Wenjing Jia

Biography

Wenjing Jia is currently a Lecturer at the School of Computing and Communications at University of Technology, Sydney (UTS). She is a core research member of iNEXT research centre – Research Centre for Innovation in IT Services and Applications.

Academic Qualifications

  • • 2007 PhD in Computing Sciences, University of Technology, Sydney
  • • 2002 Master in Communications and Information System, Fuzhou University, China
  • • 1999 Bachelor in Communications Engineering, Changchun Institute of Posts and Telecommunications (now part of the new Jilin University since 2000), China

Working Experience

  • • 02/2010–date: Lecturer, UTS
  • • 03/2007–01/2010: Chancellor’s Postdoctoral Research Fellow, UTS
  • • 09/2006–12/2006: Research Assistant (Part-Time), UTS
  • • 07/2006–03/2007: Research Associate (Part-Time), Charles Sturt University
  • • 08/1999–08/2003: Associate Lecturer, Fuzhou University, China

Professional

Professional Qualification

  • • 2012 Cisco Certified Networking Associate Instructor Trainer, Cisco Systems
  • • 2008 Cisco Networking Academy Instructor (CCAI)
  • • 2007 Cisco Certified Network Associate (CCNA), Cisco Systems
  • • from 2006 Member of IEEE (Institute of Electrical and Electronics Engineers)
  • • from 2006 Member of IAPR (International Association for Pattern Recognition)
Image of Wenjing Jia
Lecturer, School of Computing and Communications
Core Member, Centre for Innovation in IT Services Applications
PhD (UTS)
Member, The Institution of Electrical and Electronic Engineers
 
Phone
+61 2 9514 7873
Fax
+61 2 9514 4535
Room
CB10.04.442

Research Interests

Image/video analysis, computer vision, and visual pattern recognition

Can supervise: Yes

Wenjing mainly teaches Internetworking subjects for both undergraduate and postgraduate students. Subjects she has taught or is currently teaching include:

  • • LANs and Routing
  • • Routing and Internetworks
  • • WANs and VLANs

Book Chapters

Zeng, C., Jia, W., He, X.S. & Xu, M. 2013, 'Recent Advances on Graph-Based Image Segmentation Techniques' in Xiao Bai, Jian Cheng, Edwin Hancock (eds), Graph-Based Methods in Computer Vision: Developments and Applications, IGI Global, Hershey, Pennsylvania (USA), pp. 140-154.
Image segmentation techniques using graph theory has become a thriving research area in computer vision community in recent years. This chapter mainly focuses on the most up-to-date research achievements in graph-based image segmentation published in top journals and conferences in computer vision community. The representative graph-based image segmentation methods included in this chapter are classified into six categories: minimum-cut/maximum-flow model (called graph-cut in some literatures), random walk model, minimum spanning tree model, normalized cut model and isoperimetric graph partitioning. The basic rationales of these models are presented, and the image segmentation methods based on these graph-based models are discussed as the main concern of this chapter. Several performance evaluation methods for image segmentation are given. Some public databases for testing image segmentation algorithms are introduced and the future work on graph-based image segmentation is discussed at the end of this chapter.
Ye, Y., He, X.S., Li, J., Jia, W. & Wu, Q. 2009, 'Image Transformation on Hexagonal Structure Based on Conversion between 1D and 2D Coordinates' in Wen, P.; Li, Y.; Polkowski, L.; Yao, Y.; Tsumoto, S.; & Wang, G. (eds), Rough Sets and Knowledge Technology, Springer, Berlin, Germany, pp. 571-578.
View/Download from: UTSePress | Publisher's site
Spiral Architecture, a hexagonal image structure is a novel and powerful approach to machine vision system. The pixels on Spiral architecture are geometrically arranged using a 1D (Spiral) addressing scheme in an ascending order along a spiral-like curve. Spiral addition and Spiral multiplication are defined based on the Spiral addresses on Spiral Architecture. These two fundamental operations result in fast and easy translation, rotation and separation on images, and hence play very important roles for image processing on Spiral Architecture. Moreover, 2D coordinates according to rows and columns defined on Spiral Structure provide a good mapping to the ordinary 2D coordinates defined on the common square image structure. Therefore, how to convert the 1D Spiral addresses from and to the 2D coordinates on Spiral Architecture has become very important to apply the theory developed on a hexagonal image structure for image processing (e.g., rotation). In this paper, we perform a fast way to correctly locate any hexagonal pixel when its Spiral address is known, and compute the Spiral address of any hexagonal pixel when its location is known. As an illustration of the use of conversions, we demonstrate the accurate image translation and rotation using experimental results.

Conference Papers

Zeng, C., Jia, W. & He, X.S. 2013, 'Text Detection In Born-Digital Images Using Multiple Layer Images', ICASSP 2013, Vancouver Canada, May 2013 in 2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ed Li Deng, IEEE, Vancouver, Canada, pp. 1947-1951.
View/Download from: Publisher's site
In this paper, a new framework for detecting text from webpage and email images is presented. The original image is split into multiple layer images based on the maximum gradient difference (MGD) values to detect text with both strong and weak contrasts. Connected component processing and text detection are performed in each layer image. A novel texture descriptor named T-LBP, is proposed to further filter out non-text candidates with a trained SVM classifier. The ICDAR 2011 born-digital image dataset is used to evaluate and demonstrate the performance of the proposed method. Following the same performance evaluation criteria, the proposed method outperforms the winner algorithm of the ICDAR 2011 Robust Reading Competition Challenge 1.
Mudugamuwa, D.J., He, X.S. & Jia, W. 2012, 'Battle-Lemarie Wavelet Pyramid for Improved GSM Image Denoising', ICPR2012, Tsukuba, Japan, November 2012 in The 21st International Conference on Pattern Recognition (ICPR 2012), ed Del Bimbo, A.; Boyer, K. M.; Ikeuchi, K., IEEE, Tsukuba, Japan, pp. 3156-3159.
View/Download from: UTSePress
Removing noise from a digital image is a challenging problem. Application of Gaussian Scale Mixtures (GSM) in the wavelet domain has been reported to be one of the most effective denoising algorithms, published to date. The performance of this algorithm depends on the chosen wavelet representation. In this paper, we introduce an improved wavelet pyramid representation based on the Battle-Lemarie wavelet which favors the GSM denoising performance. We present the experimental denoising results using the proposed pyramid representation, and they outperform state-of-the-art GSM denoising results reported in the literature.
Mudugamuwa, D.J., He, X.S. & Jia, W. 2012, 'Efficient Super-Resolution by Finer Sub-Pixel Motion Prediction and Bilateral Filtering', Melbourne, Australia, July 2012 in 2012 IEEE International Conference on Multimedia and Expo (ICME), ed Zhang, J; n Schonfeld, D; Feng, D.D., IEEE, Melbourne, Australia, pp. 800-805.
View/Download from: UTSePress
Mudugamuwa, D.J., Jia, W., He, X.S. & Yang, J. 2011, 'An Overcomplete Pyramid Representation for Improved GSM Image Denoising', IEEE International Conference on Multimedia dan Expo, Barcelona Spain, July 2011 in Multimedia and Expo (ICME), 2011 IEEE International Conference on, ed Jin Li,Philippe Salembier,Dinei Florencio,Mohamed Hefeeda,Alexander Loui,Sethuraman (Panch) Panchanathan, IEEE Computer Society, Barcelona Spain, pp. 1-6.
Removing noise from a digital image is a challenging problem. Application of Gaussian Scale Mixtures (GSM) in the wavelet domain has been reported to be one of the most effective denoising algorithms, published to date. In this paper we investigate the impact of overcomplete wavelet image representations on the GSM image denoising algorithm. We explore the desirable local characteristics of wavelet coefficients that can enhance the efficiency of GSM denoising and based on the findings, we devise an improved over-complete pyramid representation to enhance the GSM denoising performance. We present the experimental denoising results using the proposed pyramid representation, and they outperform state-of-the-art GSM denoising results reported in the literature.
Wang, S., Wu, Q., Jia, W. & He, X.S. 2011, 'Training-Free License Plate Detection Using Vehicle Symmetry and Simple Features', Image and Vision Computing New Zealand 2011 IVCNZ, Auckland, New Zealand, November 2011 in Proceedings: Twenty-sixth International Conference Image and Vision Computing New Zealand, ed Delmas, P;Wuensche, B; James, J., Image and Vision Computing New Zealand, Auckland, New Zealand, pp. 260-265.
View/Download from: UTSePress |
In this paper, we propose a training free license plate detection method. We use a challenging benchmark dataset for license plate detection. Unlike many existing approaches, the proposed approach is a training free method, which does not require supervised training procedure and yet can achieve a reasonably good performance. Our motivation comes from the fact that, although license plates are largely variant in color, size, aspect ratio, illumination condition and so on, the rear view of vehicles is mostly symmetric with regard to the vehicle+s central axis. In addition, license plates for most vehicles are usually located on or close to the vertical axis of the vehicle body along which the vehicle is nearly symmetric. Taking advantage of such prior knowledge, the license plate detection problem is made simpler compared to the conventional scanning window approach which not only requires a large number of scanning window locations, but also requires different parameter settings such as scanning window sizes, aspect ratios and so on.
Wang, S., Wu, Q., Jia, W., He, X.S. & Yang, J. 2011, 'Learning Global and Local Features for License Plate Detection', International Conference, ICONIP, Shanghai / China, November 2011 in Neural Information Processing, ed Bao-Liang Lu, Liqing Zhang, James Kwok, Springer-Verlag, Berlin/ Heidelberg, pp. 547-556.
This paper proposes an intelligent system that is capable of automiatically detecting license plates from static images captured by a digital still camera. A supervised learning apporach is used to extract features from license plates, and both global feature and local feature are organized into a cascaded structure. In general, our framework can be divided into two stages. The first stage is constructed by extracting global correlation features and a posterior probability can be estimated to quickly determine the degree of resemblance between the evaluated image region and a license plate. The second stage is contructed by further extracting local dense-SIFT (dSIFT) features for AdaBoost supervised learning apporach and the slelected dSIFT features will be used to construct a stong classifier. Using dSIFT as a type of highly distinctive local feature, our algorithm gives high detection rate under various complex conditions. The proposed framework is compared with existing work and promising results are obtained.
Wang, S., Wu, Q., He, X.S. & Jia, W. 2011, 'More on Weak Feature: Self-correlate Histogram Distances', Pacific Rim Symposium, PSIVT, Gwangju, South Korea, November 2011 in Advances in Image and Video Technology, ed Yo-Sung Ho, Springer-Verlag, Berlin/ Heidelberg, pp. 214-223.
Wang, L., He, X.S., Du, R., Jia, W., Wu, Q. & Yeh, W. 2011, 'Facial Expression Recognition on Hexagonal Structure Using LBP-Based Histogram Variances', International Multimedia Modeling Conference, Taipei, Taiwan, January 2011 in Advances in Multimedia Modeling - Proceedings of the 17th International Multimedia Modeling Conference, MMM 2011, ed Liao HY M et al, Springer, Germany, pp. 35-45.
View/Download from: UTSePress | Publisher's site
In our earlier work, we have proposed an HVF (Histogram Variance Face) approach and proved its effectiveness for facial expression recognition. In this paper, we extend the HVF approach and present a novel approach for facial expression. We take into account the human perspective and understanding of facial expressions. For the first time, we propose to use the Local Binary Pattern (LBP) defined on the hexagonal structure to extract local, dynamic facial features from facial expression images. The dynamic LBP features are used to construct a static image, namely Hexagonal Histogram Variance Face (HHVF), for the video representing a facial expression. We show that the HHVFs representing the same facial expression (e.g., surprise, happy and sadness etc.) are similar no matter if the performers and frame rates are different. Therefore, the proposed facial recognition approach can be utilised for the dynamic expression recognition. We have tested our approach on the well-known Cohn-Kanade AU-Coded Facial Expression database. We have found the improved accuracy of HHVF-based classification compared with the HVF-based approach
Jia, W., He, X.S. & Wu, Q. 2010, 'ECCH: A Novel Color Coocurrence Histogram', Acoustics Speech and Signal Processing, Dallas, USA, March 2010 in Acoustics Speech and Signal Processing (ICASSP), 2010 IEEE International Conference, ed Wenjing Jia, Xiangjian He, Qiang Wu, IEEE Computer Society, Dallas, USA, pp. 1258-1261.
View/Download from: UTSePress | Publisher's site
In this paper, a novel color coocurrence histogram method, named eCCH which stands for color cooccurence histogram at edge points, is proposed to describe the spatial-color joint distribution of images. Unlike all existing ideas, we only investigate the color distribution of pixels located at the two sides of edge points on gradient direction lines. When measuring the similarity of two eCCHs, the Gaussian weighted histogram intersection method is adopted, where both identical and similar color pairs are considered to compensate color variations. Comparative experimental results demonstrate the performance of the proposed eCCH in terms of robustness to color variance and small computational complexity.
Mudugamuwa, D.J., Jia, W. & He, X.S. 2010, 'Asymmetric, Non-unimodal Kernel Regression for Image Processing', Digital Image Computing: Techniques and Applications, Sydney, Australia, December 2010 in Proceedings. 2010 Digital Image Computing: Techniques and Applications (DICTA 2010), ed Jian Zhang, Chunhua Shen, Glenn Geers, Qiang Wu, IEEE Computer Society, Sydney, pp. 141-145.
View/Download from: UTSePress | Publisher's site
Kernel regression has been previously proposed as a robust estimator for a wide range of image processing tasks, including image denoising, interpolation and superresolution. In this article we propose a kernel formulation that relaxes the usual symmetric and unimodal properties to effectively exploit the smoothness characteristics of natural images. The proposed method extends the kernel support along similar image characteristics to further increase the robustness of the estimates. Application of the proposed method to image denoising yields significant improvement over the previously reported regression methods and produces results comparable to the state-ofthe-art denoising techniques.
Jia, W., He, X.S. & Wu, Q. 2010, 'Segmenting Characters from License Plate Images with Little Prior Knowledge', Digital Image Computing: Techniques and Applications, Sydney, Australia, December 2010 in Proceedings. 2010 Digital Image Computing: Techniques and Applications (DICTA 2010), ed Jian Zhang, Chunhua Shen, Glenn Geers, Qiang Wu, IEEE Computer Society, Sydney, Australia, pp. 220-226.
View/Download from: UTSePress | Publisher's site
In this paper, to enable a fast and robust system for automatically recognizing license plates with various appearances, new and simple but efficient algorithms are developed to segment characters from extracted license plate images. Our goal is to segment characters properly from a license plate image region. Different from existing methods for segmenting degraded machine-printed characters, our algorithms are based on very weak assumptions and use no prior knowledge about the format of the plates, in order for them to be applicable to wider applications. Experimental results demonstrate promising efficiency and flexibility of the proposed scheme.
Zeng, C., Jia, W., He, X.S. & Yang, J. 2010, 'Graph-based text segmentation using a selected channel image', Digital Image Computing: Techniques and Applications, Sydney, Australia, December 2010 in Proceedings. 2010 Digital Image Computing: Techniques and Applications (DICTA 2010), ed Jian Zhang, Chunhua Shen, Glenn Geers, Qiang Wu, IEEE Computer Society, Sydney, Australia, pp. 535-539.
View/Download from: UTSePress | Publisher's site
This paper proposes a graph-based method for segmentation of a text image using a selected colour-channel image. The text colour information usually presents a twopolarity trend. According to the observation that the histogram distributions of the respective colour channel images are usually different from each other, we select the colour channel image with the histogram having the biggest distance between the two main peaks, which represents the main foreground colour strength and background colour strength respectively. The peak distance is estimated by the mean-shift procedure performed on each individual channel image. Then, a graph model is constructed on a selected channel image to segment the text image into foreground and background. The proposed method is tested on a public database, and its effectiveness is demonstrated by the experimental results.
Tan, Z., Jamdagni, A., He, X.S., Nanda, P., Liu, R., Jia, W. & Yeh, W. 2010, 'A Two-Tier System for Web Attack Detection Using Linear Discriminant Method', Information and Communications Security, Barcelona, Spain, December 2010 in Information and Communications Security - Lecture Notes in Computer Science 6476, ed Soriano, M; Qing, Sand; Lopez, J., Springer, Berlin Heidelberg, pp. 459-471.
View/Download from: UTSePress | Publisher's site
Computational cost is one of the major concerns of the commercial Intrusion Detection Systems (IDSs). Although these systems are proven to be promising in detecting network attacks, they need to check all the signatures to identify a suspicious attack in the worst case. This is time consuming. This paper proposes an efficient two-tier IDS, which applies a statistical signature approach and a Linear Discriminant Method (LDM) for the detection of various Web-based attacks. The two-tier system converts high-dimensional feature space into a low-dimensional feature space. It is able to reduce the computational cost and integrates groups of signatures into an identical signature. The integration of signatures reduces the cost of attack identification. The final decision is made on the integrated low-dimensional feature space. Finally, the proposed two-tier system is evaluated using DARPA 1999 IDS dataset for webbased attack detection.
He, X.S., Wei, D., Lam, K.M., Li, J., Wang, L., Jia, W. & Wu, Q. 2010, 'Canny Edge Detection Using Bilateral Filter on Real Hexagonal Structure', Advanced Concepts for Intelligent Vision System, Sydney, Australia, December 2010 in 12th International Conference - Advanced Concepts for Intelligent Vision System, ACIVS 2010, ed Paul Scheunders;Jacques Blanc-Talon; Don Bone; Wilfried Phillips; Dan Popescu, Springer-Verlag, Berlin, Germany, pp. 233-244.
View/Download from: UTSePress | Publisher's site
Edge detection plays an important role in image processing area. This paper presents a Canny edge detection method based on bilateral filtering which achieves better performance than single Gaussian filtering. In this form of filtering, both spatial closeness and intensity similarity of pixels are considered in order to prserve important visual cues provided by edges and reduce the sharpness of transitions in intensity values as well. In addition, the edge detection method proposed in this paper is achieved on sampled images represented on a real hexagon structure. Due to the compact and circular nature of hexagonal lattice, a better quality edge map is obtained on the hexagonal structure than common edge detection on square structure. Experimental results using proposed methods exhibit also the faster speed of detection on hexagonal structure.
Du, R., Wu, Q., He, X.S., Jia, W. & Wei, D. 2009, 'Facial Expression Recognition Using Histogram Variances Faces', IEEE Workshop on Applications of Computer Vision, Snowbird, USA, December 2009 in Proceedings of the Ninth IEEE Computer Society Workshop on Application of Computer Vision (WACV 2009), ed Michael, D, IEEE, USA, pp. 341-347.
View/Download from: UTSePress | Publisher's site
In human's expression recognition, the representation of expression features is essential for the recognition accuracy. In this work we propose a novel approach for extracting expression dynamic features from facial expression videos. Rather than utilising statistical models e.g. Hidden Markov Model (HMM), our approach integrates expression dynamic features into a static image, the Histogram Variances Face (HVF), by fusing histogram variances among the frames in a video. The HVFs can be automatically obtained from videos with different frame rates and immune to illumination interference. In our experiments, for the videos picturing the same facial expression, e.g., surprise, happy and sadness etc., their corresponding HVFs are similar, even though the performers and frame rates are different. Therefore the static facial recognition approaches can be utilised for the dynamic expression recognition. We have applied this approach on the well-known Cohn-Kanade AUCoded Facial Expression database then classified HVFs using PCA and Support Vector Machine (SVMs), and found the accuracy of HVFs classification is very encouraging.
He, X.S., Li, J., Wei, D., Jia, W. & Wu, Q. 2009, 'Canny edge detection on a virtual hexagonal image structure', Taipei, Taiwan, December 2009 in 2009 Joint Conferences on Pervasive Computing (JCPC2009), ed Flora Chia-I Chang et al., IEEE, Taipei, Taiwan, pp. 167-172.
View/Download from: UTSePress | Publisher's site
Canny edge detector is the most popular tool for edge detection and has many applications in the areas of image processing, multimedia and computer vision. The Canny algorithm optimizes the edge detection through noise filtering using an optimal function approximated by the first derivative of a Gaussian. It identifies the edge points by computing the gradients of light intensity function based on the fact that the edge points likely appear where the gradient magnitudes are large. Hexagonal structure is an image structure alternative to traditional square image structure. Because all the existing hardware for capturing image and for displaying image are produced based on square structure, an approach that uses linear interpolation described for conversion between square and hexagonal structures. Gaussian filtering together with gradient computation is performed on the hexagonal structure. The pixel edge strengths on the square structure are then estimated before the thresholds of Canny algorithm are applied to determine the final edge map. The experimental results show the edge detection on hexagonal structure using static and video images, and the comparison with the results using Canny algorithm on square structure.
He, X.S., Zheng, L., Wu, Q., Jia, W., Samali, B. & Palaniswami, M.S. 2008, 'Segmentation of Characters on Car License Plates', International Workshop on Multimedia Signal Processing, Cairns, Australia, October 2008 in IEEE 10th Workshop on Multimedia Signal Processing, 2008, ed Feng D. and et al., IEEE, USA, pp. 399-402.
View/Download from: UTSePress | Publisher's site
License plate recognition usually contains three steps, namely license plate detection/localization, character segmentation and character recognition. When reading characters on a license plate one by one after license plate detection step, it is crucial to accurately segment the characters. The segmentation step may be affected by many factors such as license plate boundaries (frames). The recognition accuracy will be significantly reduced if the characters are not properly segmented. This paper presents an efficient algorithm for character segmentation on a license plate. The algorithm follows the step that detects the license plates using an AdaBoost algorithm. It is based on an efficient and accurate skew and slant correction of license plates, and works together with boundary (frame) removal of license plates. The algorithm is efficient and can be applied in real-time applications. The experiments are performed to show the accuracy of segmentation.
He, X.S., Jia, W. & Wu, Q. 2008, 'An Approach of Canny Edge Detection with Virtual Hexagonal Image Structure', International Conference on Control, Automation, Robotics and Vision, Hanoi, Vietnam, December 2008 in International Conference on Control, Automation, Robotics and Vision, ed Changyun Wen and Thuong Cat Pham, IEEE, Singapore, pp. 879-882.
View/Download from: UTSePress | Publisher's site
Edge detection plays an important role in the areas of image processing, multimedia and computer vision. Gradient-based edge detection is a straightforward method to identify the edge points in the original grey-level image. It is intuitive that, in the human vision system, the edge points always appear where the gradient magnitude assumes a maximum. Hexagonal structure is an image structure alternative to traditional square image structure. The geometrical arrangement of pixels on a hexagonal structure can be described as a collection of hexagonal pixels. Because all the existing hardware for capturing image and for displaying image are produced based on square structure, an approach that uses bilinear interpolation and tri-linear interpolation is applied for conversion between square and hexagonal structures. Based on this approach, an edge detection method is proposed. This method performs Gaussian filtering to suppress image noise and computes gradients on the hexagonal structure. The pixel edge strengths on the square structure are then estimated before Canny's edge detector is applied to determine the final edge map. The experimental results show that the proposed method improves the edge detection accuracy and efficiency.
Chen, Y., Wu, Q., He, X.S., Jia, W. & Hintz, T.B. 2008, 'A Modified Mahalanobis Distance for Human Detection in Out-door Environments', U-Media, Lanzhou, China, July 2008 in First IEEE International Conference on Ubi-media Computing (U-Media 2008), ed Lian Li et al., IEEE, Beijing, China, pp. 243-248.
View/Download from: UTSePress | Publisher's site
This paper proposes a novel method for human detection from static images based on pixel structure of input images. Each image is divided into four parts, and a weight is assigned to each part of the image. In training stage, all sample images including human images and non-human images are used to construct a Mahalanobis distance map through statistically analyzing the difference between the different blocks on each original image. A projection matrix will be created with Linear Discriminant Method (LDM) based on the Mahalanobis distance map. This projection matrix will be used to transform multi-dimensional feature vectors into one dimensional feature domain according to a pre-calculated threshold to distinguish human figures from non-human figures. In comparison with the method without introducing weights, the proposed method performs much better. Encouraging experimental results have been obtained based on MIT dataset and our own dataset.
He, X.S., Zheng, L., Wu, Q., Jia, W., Samali, B. & Palaniswami, M.S. 2008, 'A hierarchically combined classifier for license plate recognition', IEEE International Conference on Computer and Information Technology, Sydney, July 2008 in IEEE 8th International Conference on Computer and Information Technology (CIT2008), ed Qiang Wu et al., IEEE Computer Society, USA, pp. 372-377.
View/Download from: UTSePress | Publisher's site
High accuracy and fast recognition speed are two requirements for real-time and automatic license plate recognition system. In this paper, we propose a hierarchically combined classifier based on an inductive learning based method and an SVM-based classification. This approach employs the inductive learning based method to roughly divide all classes into smaller groups. Then the SVM method is used for character classification in individual groups. Both start from a collection of samples of characters from license plates. After a training process using some known samples in advance, the inductive learning rules are extracted for rough classification and the parameters used for SVM-based classification are obtained. Then, a classification tree is constructed for further fast training and testing processes for SVM-based classification. Experimental results for the proposed approach are given. From the experimental results, we can make the conclusion that the hierarchically combined classifier is better than either the inductive learning based classification or the SVM-based classification in terms of error rates and processing speeds.
He, X.S., Zhang, H., Jia, W., Wu, Q. & Hintz, T.B. 2007, 'Combining Global and Local Features for Detection of License Plates in Video', Image and Vision Computing Conference, Hamilton, New Zealand, December 2007 in Proceedings of Image and Vision Computing New Zealand 2007, ed Cree, Michael J., Image and Vision Computing New Zealand, Hamilton, New Zealand, pp. 288-293.
View/Download from: UTSePress |
Jia, W., He, X.S., Zhang, H. & Wu, Q. 2007, 'Combining Edge and Colour Information for Number Plate Detection', Image and Vision Computing Conference, Hamilton, New Zealand, December 2007 in Proceedings of Image and Vision Computing New Zealand 2007, ed Cree, Michael J., Image and Vision Computing New Zealand, Hamilton, New Zealand, pp. 227-232.
View/Download from: UTSePress
This paper presents a method for vehicle number plate detection which combines edge and colour features of number plates. We concentrate on two key issues of this application: speed and robustness. Our focus is put on detecting parts of a number plate, instead of the number plate itself as a whole. To achieve the target of real-time detection, two simple features based on a rede-ned vertical edge map are constructed. To address the illumination-sensitive problem of using colour information, a Gaussian weighted histogram intersection (GWHI) method is proposed. The above new ideas compose the major part of the algorithm. Our experimental results demonstrate a promising preliminary result on detecting yellow number plates in terms of detection speed and robustness, which shows the feasibility of the proposed method.
He, X.S., Li, J., Jia, W., Wu, Q. & Hintz, T.B. 2007, 'Local Binary Patterns on Hexagonal Image Structure', IEEE International Conference on Computer and Information Technology, Aizu-Wakamatsu City, Fukushima, Japan, October 2007 in Proceedings of 7th IEEE International Conference on Computer and Information Technology, ed Miyazaki, T.; Paik, I.; Wei, D., IEEE, USA, pp. 639-644.
View/Download from: UTSePress | Publisher's site
Local binary pattern (LBP) was designed and widely used for efficient texture classification. It has been used for face recognition and has potential applications in many other research areas such as human detection. LBP provides a simple and effective way to represent patterns. Uniform LBPs play an important role for LBP-based pattern /object recognition as they include majority of LBPs. In this paper, we present LBP codes on hexagonal image structure. We show that LBPs defined on hexagonal structure have higher percentages of uniform LBPs that will lead to a more efficient and accurate recognition scheme for image classification.
He, X.S., Li, J., Chen, Y., Wu, Q. & Jia, W. 2007, 'Local Binary Patterns for Human Detection on Hexagonal Structure', IEEE International Symposium on Multimedia, Taichung, Taiwan, December 2007 in Proceedings of the 2007 IEEE International Symposium on Multimedia (ISM-07), ed Bulterman, Dick; Mori, Kinji; Tsai, Jeffrey J. P. et al, IEEE Computer Society, Los Alamitos, California, pp. 65-71.
View/Download from: UTSePress | Publisher's site
Local Binary Pattern (LBP) was designed and has been widely used for efficient texture classification. LBP provides a simple and effective way to represent texture patterns. Uniform LBPs play an important role for LBP-based pattern/object recognition as they include majority of LBPs. On the other hand, Human detection based on Mahalanobis Distance Map (MDM) recognizes appearance of human based on geometrical structure. Each MDM shows a clear texture pattern that can be classified using LBPs. In this paper, we compute LBPs of MDMs on a hexagonal structure. The circular pixel arrangement in hexagonal structure results in higher accuracy for LBP representation than on square structure. Chi-square as a measure is used for human detection based on uniform LBPs obtained. We show that our method using LBPs built on MDMs has a higher human detection rate and a lower false positive rate compared to the method merely based on MDMs. We will also show using experimental results that LBPs on hexagonal structure lead to more robust human classification.
Jia, W., Tien, D., He, X.S., Hope, B.A. & Wu, Q. 2007, 'Applying Local Cooccurring Patterns for Object Detection from Aerial Images', International Conference on Visual Information Systems, Shanghai, China, June 2007 in International Conference on Visual Information Systems - Lecture Notes in Computer Science, ed NA, Springer, Berlin / Heidelberg, pp. 478-489.
View/Download from: UTSePress | Publisher's site
Developing a spatial searching tool to enhance the search capabilities of large spatial repositories for Geographical Information System (GIS) update has attracted more and more attention. Typically, objects to be detected are represented by many local features or local parts. Testing images are processed by extracting local features which are then matched with the object++s model image. Most existing work that uses local features assumes that each of the local features is independent to each other. However, in many cases, this is not true. In this paper, a method of applying the local cooccurring patterns to disclose the cooccurring relationships between local features for object detection is presented. Features including colour features and edge-based shape features of the interested object are collected. To reveal the cooccurring patterns among multiple local features, a colour cooccurrence histogram is constructed and used to search objects of interest from target images. The method is demonstrated in detecting swimming pools from aerial images. Our experimental results show the feasibility of using this method for effectively reducing the labour work in finding man-made objects of interest from aerial images.
He, X.S., Jia, W., Wu, Q. & Hintz, T.B. 2007, 'Parallel Edge Detection on a Virtual Hexagonal Structure', International Conference on Grid and Pervasive Computing, Paris, France, May 2007 in International Conference on Grid and Pervasive Computing - Lecture Notes in Computer Science, ed NA, Springer, Berlin / Heidelberg, pp. 751-756.
View/Download from: UTSePress | Publisher's site
This paper presents an edge detection method based on bilateral filtering taking into account both spatial closeness and intensity similarity of pixels in order to preserve important visual cues provided by edges and reduce the sharpness of transitions in intensity values as well. In addition, the edge detection method proposed in this paper is achieved on sampled images represented on a newly developed virtual hexagonal structure. Due to the compact and circular nature of the hexagonal lattice, a better quality edge map is obtained. We also present a parallel implementation for edge detection on the virtual hexagonal structure that significantly increases the computation speed.
He, X.S., Hintz, T.B., Li, J., Zhang, H., Wu, Q. & Jia, W. 2007, 'Local Binary Pattern on Hexagonal Structure for Face Matching', International Conference on Image Processing, Computer Vision and Pattern Recognition, Las Vegas, USA, June 2007 in Proceedings of the 2007 International Conference on Image Processing, Computer Vision and Pattern Recognition, ed Arabnia, Hamid R., CSREA Press, USA, pp. 455-460.
View/Download from: UTSePress
He, X.S., Jia, W., Li, J., Wu, Q. & Hintz, T.B. 2007, 'An Approach to Edge Detection on a Virtual Hexagonal Structure', Digital Image Computing Techniques and Applications, Glenelg, Australia, December 2007 in Digital Image Computing Techniques and Applications, ed Bottema, Murk et al., IEEE Computer Society, Los Alamitos, USA, pp. 340-345.
View/Download from: UTSePress | Publisher's site
Hexagonal structure is another image structure alternative to traditional square image structure for image processing and computer vision. The geometrical arrangement of pixels on a hexagonal structure can be described as a collection of hexagonal pixels. Because all the existing hardware for capturing image and for displaying image are produced based on square structure, it becomes important to find a proper software approach to mimic hexagonal structure so that images represented on the traditional square structure can be smoothly converted from or to the images on hexagonal structure. For accurate image processing, it is critical to best maintain the image resolution during the image conversion. In this paper, a bilinear interpolation algorithm that is used to convert an image from square structure to hexagonal structure is presented. Based on this, an edge detection method is proposed. Our experimental results show that the bilinear interpolation improves the edge detection accuracy.
Chen, Y., Wu, Q., He, X.S., Jia, W. & Hintz, T.B. 2007, 'Pixel Structure Based on Hausdorff Distance for Human Detection in Outdoor Environments', Digital Image Computing Techniques and Applications, Glenelg, Australia, December 2007 in Digital Image Computing Techniques and Applications, ed Bottema, Murk et al., IEEE Computer Society, Los Alamitos, USA, pp. 67-72.
View/Download from: UTSePress | Publisher's site
This paper proposes a novel method for human detection from static images based on pixel structure of input images. In training stage, all sample images consisting of human images and non-human images are used to construct a Hausdorff distance map based on statistically analyzing the difference between the different blocks on each original image. A projection matrix will be created with Linear Discriminant Method (LDM) based on the Hausdorff distance map. This projection matrix will be used to transform multidimensional feature vectors (distance maps of testing images) into a feature in a one-dimensional domain. The decision will be made on the simple one dimensional feature domain according to a precalculated threshold to distinguish human figures from non-human figures. In comparison with the common method based on Mahalanobis distance maps, the proposed method based on Hausdorff distance maps performs much better. Encouraging experimental results have been obtained using 800 human images and 800 non-human images.
Jia, W. & Tien, D. 2007, 'Discovering Local Cooccurring Patterns from Aerial Images', International Conference on Information Technology and Applications, Harbin, China, January 2007 in Proceedings of the 4th International Conference on Information Technology and Applications, ed Tien, D; Shi, G; Wang, G., Macquarie Scientific Publishing, Sydney, Australia, Sydney, Australia, pp. 300-305.
View/Download from: UTSePress
Tien, D. & Jia, W. 2007, 'Automatic Road Extraction from Aerial Images: A Contemporary Survey', International Conference on Information Technology and Applications, Harbin, China, January 2007 in Proceedings of the 4th International Conference on Information Technology and Applications, ed Tien, D; Shi, G; Wang, G., Macquarie Scientific Publishing, Sydney, Australia, Sydney, Australia, pp. 294-299.
View/Download from: UTSePress
Jia, W., Zhang, H., He, X.S. & Wu, Q. 2006, 'Symmetric color ratio in spiral architecture', Asian Conference on Computer Vision, Hyderabad, India, January 2006 in Computer Vision - ACCV 2006, Pt ii, Lecture Notes in Computer Science, ed NA, Springer-Verlag Berlin, Berlin, Germany, pp. 204-213.
View/Download from: UTSePress | Publisher's site
Color ratio gradient (CRG) is a robust method used for color image retrieval and object recognition. It has been proven to be illumination-independent and geometry-insensitive when tested on scenery images. However, the color ratio gradient produces unsa
Jia, W., Zhang, H., He, X.S. & Wu, Q. 2006, 'Gaussian weighted histogram intersection for license plate classification', International Conference on Pattern Recognition, Hong Kong, PEOPLES R CHINA, August 2006 in 18Th International Conference On Pattern Recognition, Vol 3, Proceedings, ed Tang, YY; Wang, SP; Lorette, G; Yeung, DS; Yan, H, IEEE Computer Soc, Los Alamitos, USA, pp. 574-577.
View/Download from: UTSePress | Publisher's site
The conventional histogram intersection (HI) algorithm computes the intersected section of the corresponding color histograms in order to measure the matching rate between two color images. Since this algorithm is strictly based on the matching between b
Wu, Q., Zhang, H., Jia, W., He, X.S., Yang, J. & Hintz, T.B. 2006, 'Car plate detection using cascaded tree-style learner based on hybrid object features', Advanced Video and Signal Based Surveillance, Sydney, Australia, November 2006 in Proceedings of international conference on video and signal based surveillance 2006, ed Piccardi, M; Hintz, T; Pavlidis, I; Regazzoni, C; He, X, IEEE, Sydney, Australia, pp. 1-6.
View/Download from: UTSePress
Jia, W., Zhang, H., He, X.S. & Wu, Q. 2006, 'Image matching using color edge coocurrence histogram', IEEE Conference on Systems, Man and Cybernetics, Taipei, Taiwan, October 2006 in Proceedings of the 2006 IEEE international conference on systems, man and cybernetics, ed Lee, T; Zhou, M, IEEE, Taiwan, pp. 2413-2419.
View/Download from: UTSePress | Publisher's site
In this paper, a novel colour edge cooccurrence histogram (CECH) method is proposed to match images by measuring similarities between their CECH histograms. Unlike the previous colour edge cooccurrence histogram proposed by Crandall and Luo (2004 ) we only investigate those pixels which are located at the two sides of edge points in their gradient direction lines and at a distance away from the edge points. When measuring similarities between two CECH histograms, a newly proposed Gaussian weighted histogram intersection (GWHI) method is extended for this purpose. Both identical colour pairs and similar colour pairs are taken into account in our algorithm, and the weights are decided by the larger distance between two colour pairs involved in matching. The proposed algorithm is tested for matching vehicle number plate images captured under various illumination conditions. Experimental results demonstrate that the proposed algorithm can be used to compare images in real-time, and is robust to illumination variations and insensitive to the model images selected.
Jia, W., He, X.S. & Tien, D. 2006, 'Automatically detecting road sign text from natural scene video', IEEE Tencon (IEEE Region 10 Conference), Hong Kiong, November 2006 in Proceedings of IEEE region 10 conference 2006, ed Luk, K; Lau, D; Lau, R, IEEE, Hong Kong, pp. 1-4.
View/Download from: UTSePress
He, X.S., Hintz, T.B., Wu, Q., Wang, H. & Jia, W. 2006, 'A new simulation of spiral architecture', International Conference on Image Processing, Computer Vision and Pattern Recognition, Las Vegas, USA, June 2006 in IPCU 06 procedings, ed Arbnia, H; He, X; Hintz, T; Liu, D; Sim, K, CSREA Press, Las Vegas USA, pp. 570-575.
View/Download from: UTSePress
He, X.S., Wang, H., Hur, N., Jia, W., Wu, Q., Kim, J.C. & Hintz, T.B. 2006, 'Uniformly partitioning images on a virtual hexagonal structure', International Conference on Control, Automation, Robotics and Vision, Singapore, December 2006 in 2006 8th International conference on control automation robotics and vision (ICARCV 2006), ed Xie, L; Cheah, C, IEEE, Singapore, pp. 891-896.
View/Download from: UTSePress | Publisher's site
Hexagonal structure is different from the traditional square structure for image representation. The geometrical arrangement of pixels on hexagonal structure can be described in terms of a hexagonal grid. Uniformly separating image into seven similar copies with a smaller scale has commonly been used for parallel and accurate image processing on hexagonal structure. However, all the existing hardware for capturing image and for displaying image are produced based on square architecture. It has become a serious problem affecting the advanced research based on hexagonal structure. Furthermore, the current techniques used for uniform separation of images on hexagonal structure do not coincide with the rectangular shape of images. This has been an obstacle in the use of hexagonal structure for image processing. In this paper, we briefly review a newly developed virtual hexagonal structure that is scalable. Based on this virtual structure, algorithms for uniform image separation are presented. The virtual hexagonal structure retains image resolution during the process of image separation, and does not introduce distortion. Furthermore, images can be smoothly and easily transferred between the traditional square structure and the hexagonal structure while the image shape is kept in rectangle
Jia, W., Zhang, H., He, X.S. & Wu, Q. 2006, 'Refined gaussian weighted histogram intersection and its application in number plate categorization', International Conference Computer Graphics, Imaging and Visualization, Sydney Australia, July 2006 in Proceedings, computer graphics, imaging and visualisation, ed Banissi, M; Huang, M; Qu, Q, IEEE Computer Society, Los Alamitos, USA, pp. 249-254.
View/Download from: UTSePress | Publisher's site
This paper proposes a refined Gaussian weighted histogram intersection for content-based image matching and applies the method for number plate categorization. Number plate images are classified into two groups based on their colour similarities with the model image of each group. The similarities of images are measured by the matching rates between their colour histograms. Histogram intersection (HI) is used to calculate the matching rates of histograms. Since the conventional histogram intersection algorithm is strictly based on the matching between bins of identical colours, the final matching rate could easily be affected by colour variation caused by various environment changes. In our recent paper [9], a Gaussian weighted histogram intersection (GWHI) algorithm has been proposed to facilitate the histogram matching via taking into account matching of both identical colours and similar colours. The weight is determined by the distance between two colours. When applied to number plate categorization, the GWHI algorithm demonstrates to be more robust to colour variations and produces a classification with much lower intra-class distance and much higher interclass distance than previous HI algorithms. However, the processing speed of this GWHI method is still not satisfying. In this paper, the GWHI method is further refined, where a colour quantization method is utilized to reduce the number of colours without introducing apparent perceptual colour distortion. New experimental results demonstrate that using the refined GWHI method, image categorization can be done more efficiently.
Zhang, H., Jia, W., He, X.S. & Wu, Q. 2006, 'A fast algorithm for license plate detection in various conditions', IEEE Conference on Systems, Man and Cybernetics, Taibei, China, November 2006 in Proceedings of the 2006 IEEE International Conference on System, Man and Cybernetics, ed Lee, T; Zhou, M, IEEE, Los Alamitos, USA, pp. 2420-2425.
View/Download from: UTSePress | Publisher's site
This paper proposes a fast algorithm detecting license plates in various conditions. There are three main contributions in this paper. The first contribution is that we define a new vertical edge map, with which the license plate detection algorithm is extremely fast. The second contribution is that we construct a cascade classifier which is composed of two kinds of classifiers. The classifiers based on statistical features decrease the complexity of the system. They are followed by the classifiers based on Haar-features, which make it possible to detect license plate in various conditions. Our algorithm is robust to the variance of the illumination, view angle, the position, size and color of the license plates when working in complex environment. The third contribution is that we experimentally analyze the relations of the scaling factor with detection rate and processing time. On the basis of the analysis, we select the optimal scaling factor in our algorithm. In the experiments, both high detection rate (with low false positive rate) and high speed are achieved when the algorithm is used to detect license plates in various complex conditions.
He, X.S., Jia, W., Wu, Q., Hur, N., Hintz, T.B., Wang, H. & Kim, J.C. 2006, 'Basic transformation on virtual hexagonal structure', International Conference Computer Graphics, Imaging and Visualization, Sydney, Australia, July 2006 in Proceedings. 2006 international conference on computer graphics, imaging and visualisation, ed Banssi, E; Sarfraz, M; Huang, M; Wu, Q, IEEE Computer society, Los Alamitos, USA, pp. 243-248.
View/Download from: UTSePress | Publisher's site
Hexagonal structure is different from the traditional square structure for image representation. The geometrical arrangement of pixels on hexagonal structure can be described in terms of a hexagonal grid. Hexagonal structure provides an easy way for image translation and rotation transformations. However, all the existing hardware for capturing image and for displaying image are produced based on square architecture. It has become a serious problem affecting the advanced research based on hexagonal structure. In this paper, we introduce a new virtual hexagonal structure. Based on this virtual structure, a more flexible and powerful image translation and rotation are performed. The virtual hexagonal structure retains image resolution during the process of image transformations, and does not introduce distortion. Furthermore, images can be smoothly and easily transferred between the traditional square structure and the hexagonal structure.
He, X.S., Jia, W., Hur, N., Wu, Q. & Kim, J.C. 2006, 'Image translation and rotation on hexagonal structure', IEEE International Conference on Computer and Information Technology, Seoul, Korea, September 2006 in Sixth IEEE International conference on computer and information technology, ed Euh, Y-D, IEEE CS, Seoul, Korea, pp. 1-6.
View/Download from: UTSePress | Publisher's site
Image translation and rotation are becoming essential operations in many application areas such as image processing, computer graphics and pattern recognition. Conventional translation moves image from pixels to pixels and conventional rotation usually comprises of computation-intensive CORDIC operations. Traditionally, images are represented on a square pixel structure. In this paper, we perform reversible and fast image translation and rotation based on a hexagonal structure. An image represented on the hexagonal structure is a collection of hexagonal pixels of equal size. The hexagonal structure provides a more flexible and efficient way to perform image translation and rotation without losing image information. As there is not yet any available hardware for capturing image and for displaying image on a hexagonal structure, we apply a newly developed virtual hexagonal structure. The virtual hexagonal structure retains image resolution during the process of image transformations, and almost does not introduce distortion. Furthermore, images can be smoothly and easily transferred between the traditional square structure and the hexagonal structure.
Jia, W., Zhang, H., He, X.S. & Wu, Q. 2006, 'A comparison on histogram based image matching methods', Advanced Video and Signal Based Surveillance, Sydney, Australia, November 2006 in Proceedings of the IEEE international conference on video and signal based surveillance, ed Piccardi, M; Hintz, T; Pavlidis, I; Regazzoni, G; He, X, IEEE Computer Society, Los Alamitos, USA, pp. 1-6.
View/Download from: UTSePress
Zhang, H., Jia, W., He, X.S. & Wu, Q. 2006, 'Learning based license plate detection using global and local features', International Conference on Pattern Recognition, Hong Kong, August 2006 in 18th International Conference on Pattern Recognition, ed Tang, Y; Wang, P; Lorette, G, Yeong, D, IEEE Computer Society, Los Akamitos USA, pp. 1102-1105.
View/Download from: UTSePress | Publisher's site
This paper proposes a license plate detection algorithm using both global statistical features and local Haar-like features. Classifiers using global statistical features are constructed firstly through simple learning procedures. Using these classifiers, more than 70% of background area can be excluded from further training or detecting. Then the AdaBoost learning algorithm is used to build up the other classifiers based on selected local Haar-like features. Combining the classifiers using the global features and the local features, we obtain a cascade classifier. The classifiers based on global features decrease the complexity of the system. They are followed by the classifiers based on local Haar-like features, which makes the final classifier invariant to the brightness, color, size and position of license plates. The encouraging detection rate is achieved in the experiments
Zhang, H., Jia, W., He, X.S. & Wu, Q. 2006, 'Real-time license plate detection under various conditions', International Conference on Ubiquitous and Intelligence Computing, Wuhan, China, September 2006 in Third International Conference, International Conference on Ubiquitous and Intelligence Computing 2006 - Lecture Notes in Computer Sciences, ed NA, Springer, Germany, pp. 192-199.
View/Download from: UTSePress | Publisher's site
This paper proposes an algorithm for real-time license plate detection. In this algorithm, the relatively easy car plate features are adopted including the simple statistical feature and Harr-like feature. The simplicity of the object features used is very helpful to real-time processing. The classifiers based on statistical features decrease the complexity of the system. They are followed by the classifiers based on Haar-like features, which makes the final classifier invariant to the brightness, color, size and position of license plates. The experimental results obtained by the proposed algorithm exhibit the encouraging performance.
He, X.S., Jia, W., Hur, N., Wu, Q., Kim, J. & Hintz, T.B. 2006, 'Bi-lateral edge detection on a virtual hexagonal structure', International Symposium on Visual Computing, United States, November 2006 in International Symposium on Visual Computing 2006 - Lecture notes in computing science, ed NA, Springer, New York USA, pp. 176-185.
View/Download from: UTSePress
Jia, W., Zhang, H. & He, X.S. 2005, 'Mean shift for accurate number plate detection', International Conference on Information Technology and Applications, Sydney, Australia, July 2005 in Proceedings of Third International Conference On Information Technology And Applications, Vol 1, ed He, X; Hintz, T; Piccardi, M; Wu, Q; Huang, M; Tien, D, IEEE, Sydney, Australia, pp. 732-737.
View/Download from: UTSePress | Publisher's site
This paper presents a robust method for number plate detection, where mean shift segmentation is used to segment color vehicle images into candidate regions. Three features are extracted in order to decide whether a candidate region contains a number pla
Zhang, H., Jia, W., He, X.S. & Wu, Q. 2005, 'Modified Colour Ratio Gradients', International Workshop on Multimedia Signal Processing, Shanghai, China, October 2005 in 2005 IEEE Seventh workshop on Multimedia Signal Processing, ed Zhang, X; Sorenson, J; Wu, Q; Shi, Y; Ostermann, J; Manj H; Goldgot, D;, IEEE, NJ, USA, pp. 317-320.
View/Download from: UTSePress | Publisher's site
Color ratio gradient is an efficient method used for color image retrieval and object recognition, which is shown to be illumination-independent and geometry-insensitive when tested on scenery images. However, color ratio gradient produces unsatisfied matching result while dealing with relatively uniform objects without rich color texture. In addition, performance of color ratio gradient degenerates while processing unsaturated color image objects. In this paper, a scheme with modified color ratio gradient is presented, which addresses the two problems above. Experimental results using the proposed method in this paper exhibit more robust performance
He, X.S. & Jia, W. 2005, 'Hexagonal structure for intelligent vision', International Conference on Information and Communications Technology, Karachi, Pakistan, August 2005 in Proceedings of IBS ICICT 2005 1st International conference on information and communication technologies, ed Khan, W.A; Ahmed, F;, IEEE, Piscataway, USA, pp. 52-64.
View/Download from: UTSePress |
Using hexagonal grids to represent digital images have been studied for more than 40 years. Increased processing capabilities of graphic devices and recent improvements in CCD technology have made hexagonal sampling attractive for practical applications and brought new interests on this topic. The hexagonal structure is considered to be preferable to the rectangular structure due to its higher sampling efficiency, consistent connectivity and higher angular resolution and is even proved to be superior to square structure in many applications. Since there is no mature hardware for hexagonal-based image capture and display, square to hexagonal image conversion has to be done before hexagonal-based image processing. Although hexagonal image representation and storage has not yet come to a standard, experiments based on existing hexagonal coordinate systems have never ceased. In this paper, we firstly introduced general reasons that hexagonally sampled images are chosen for research. Then, typical hexagonal coordinates and addressing schemes, as well as hexagonal based image processing and applications, are fully reviewed.
Jia, W., Zhang, H., He, X.S. & Piccardi, M. 2005, 'Mean shift for accurate license plate localisation', International Conference on Intelligent Transportation Systems, Vienna, Austria, September 2005 in ITSC '05 - 8th International Conference on Intelligent Transportation Systems, ed Pfliegl, R., IEEE, Vienna, Austria, pp. 566-571.
View/Download from: UTSePress
Jia, W., He, X.S. & Piccardi, M. 2004, 'Automatic License Plate Recognition: A review', International Conference Imaging Science, Systems and Technology, Las Vegas, Nevada, USA, June 2004 in Proceedings of the International Conference on Imaging Science, Systems and Technology, ed Arabnia Hamrid R., CSREA Press, Las Vegas, Nevada, USA, pp. 43-48.
View/Download from: UTSePress | Publisher's site
Jia, W., He, X.S. & Lin, Q. 2004, 'Echocardiography Sequential Images Compression Based on Region of Interest', International Conference on Information Technology and Applications, Harbin, China, January 2004 in Proceedings of 2nd International Conference on Information Technology and Applications, ed Shi, G., Macquarie Scientific Publishing, Sydney, Australia, pp. 232-237.
View/Download from: UTSePress | Publisher's site
Jia, W., He, X.S. & Wu, Q. 2004, 'Edge Analysis on Rectangular and Hexagonal Architectures', International Conference on Information and Communication Technologies, Bangkok, Thailand, November 2004 in Proceedings of the International Conference on Information and Communication Technologies, ed Batavski D A. and Fedoseev, S A., Assumption University, Bangkok, Thailand, pp. 69-75.
View/Download from: UTSePress

Journal Articles

Zeng, C., Jia, W. & He, X.S. 2012, 'An Algorithm for Colour-based Natural Scene Text Segmentation', Lecture Notes in Computer Science, vol. 7139, pp. 58-68.
View/Download from: UTSePress
Before the step for text recognition, a text image needs to be segmented into foreground containing only the text area and background. In this paper, a method is proposed for segmenting colour natural scene texts which suffer from a wide range of degradations with complex background. A text image is firstly processed by two 3-means clustering operations with different distance measurements. Then, a modified connected component (CC)-based validation method is used to obtain the text area after clustering. Thirdly, a proposed objective segmentation evaluation method is utilised to choose the final segmentation result from the two segmented text images. The proposed method is compared with other existing methods based on the ICDAR2003 public database. Experimental results show the effectiveness of the proposed method.
Han, L., Fu, C., Zou, D., Lee, C. & Jia, W. 2012, 'Task-based behavior detection of illegal codes', Mathematical and Computer Modelling, vol. 55, no. 1-2, pp. 80-86.
View/Download from: UTSePress | Publisher's site
Detecting unseen illegal codes is always a challenging task. As the main action to deal with this problem, the behavior detection is unsatisfactory in both effectiveness and efficiency. This paper proposes task-based behavior detection (TBBD) which detects new illegal codes based on the user++s task instead of only on the software behavior. First, the paper proposes three prerequisites of TBBD and four judgment rules, i.e., resource abnormal rule, relation abnormal rule, space abnormal rule and time abnormal rule. Then, by analyzing the effectiveness and comparison of the four judgment rules, we present an explicit judgment process of TBBD. Finally, the paper carries on the experiments. The test result verifies the validity and feasibility of TBBD.
Jiang, Y., He, X.S., Lin, F. & Jia, W. 2011, 'An Encoding and Labeling Scheme Based on Continued Fraction for Dynamic XML', Journal of Software, vol. 6, no. 10, pp. 2043-2049.
Much research about labeling schemes has been conducted to efficiently determine the ancestor-descendant relationships and the document-order between any two random XML nodes without re-labeling for updates. In this paper, we present an efficient XML encoding and labeling scheme for dynamic XML document, named Continued Fraction-based Encoding (CFE). The proposed CFE scheme labels nodes with continued fractions and has the following three important properties: (1) CFE codes can be inserted between any two consecutive CFE codes with the orders kept and without re-encoding the existing nodes; (2) CFE is orthogonal to specific labeling schemes, thus it can be applied broadly to different labeling schemes or other applications to efficiently process the updates; (3) CFE supports all structural relationships query in XPath. Two test data sets were built for evaluation. The experimental results show that CFE provides fairly reasonable XML query processing performance while completely avoiding relabeling for updates.
He, X.S., Wu, Q., Jia, W. & Hintz, T.B. 2008, 'Edge Detection on Hexagonal Structure', Journal of Algorithms & Computational Technoiogy, vol. 2, no. 1, pp. 61-78.
View/Download from: UTSePress
Gradient-based edge detection is a straightforward method to identify the edge points in the original grey-level image. It is intuitive that in the human vision system the edge points always appear where the grey-level value is greatly changed. Spiral Architecture is a relatively new image data structure that is inspired from anatomical considerations of the primate's vision. In Spiral Architecture, each image is represented as a collection of hexagonal pixels. Edge detection on Spiral Architecture has features of fast computation and accurate localization. In this paper, we present and compare gradient-based edge detection algorithms on Spiral Architecture. The experimental results show that the edge detection on Spiral Architecture outperforms that on traditional square image structure.
He, X.S., Wu, Q., Hintz, T.B. & Jia, W. 2008, 'Gradient-Based Edge Detection on a Hexagonal Structure', Business Review, vol. 3, no. 1, pp. 133-143.
View/Download from: UTSePress
Gradient-based edge detection is a straightforward method to identify the edge points in the original grey-level image. It is intuitive that in the human vision system the edge points always appear where the grey-level value is greatly changed. Spiral Architecture is a relatively new image data structure that is inspired from anatomical considerations of the primate's vision. In Spiral Architecture, each image is represented as a collection of hexagonal pixels. Edge detection on Spiral Architecture has features of fast computation and accurate localization. In this paper, we present and compare gradient-based edge detection algorithms on Spiral Architecture. The experimental results show that the edge detection on Spiral Architecture outperforms that on traditional square image structure.
Jia, W., Zhang, H. & He, X.S. 2007, 'Region-based License Plate Detection', Journal Of Network And Computer Applications, vol. 30, no. 4, pp. 1324-1333.
View/Download from: UTSePress | Publisher's site
Automatic license plate recognition (ALPR) is one of the most important aspects of applying computer techniques towards intelligent transportation systems. In order to recognize a license plate efficiently, however, the location of the license plate, in most cases, must be detected in the first place. Due to this reason, detecting the accurate location of a license plate from a vehicle image is considered to be the most crucial step of an ALPR system, which greatly affects the recognition rate and speed of the whole system. In this paper, a region-based license plate detection method is proposed. In this method, firstly, mean shift is used to filter and segment a color vehicle image in order to get candidate regions. These candidate regions are then analyzed and classified in order to decide whether a candidate region contains a license plate. Unlike other existing license plate detection methods, the proposed method focuses on regions, which demonstrates to be more robust to interference characters and more accurate when compared with other methods.
Zhang, H., Jia, W., He, X.S. & Wu, Q. 2007, 'Learning-Based License Plate Detection in Vehicle Image Database', International Journal of Intelligent Information and Database Systems, vol. 1, no. 2, pp. 228-243.
View/Download from: UTSePress | Publisher's site
This paper proposes a learning-based algorithm to detect license plates of vehicles from vehicle image database. There are three main contributions in this paper. The first contribution is to define a novel vertical edge map, which makes the image processing more effectively. The second contribution is to propose a learning-based cascade classifier composing of two kinds of sub-classifiers, which makes the system very robust. The third contribution is to experimentally estimate the parameter of scaling factor and chose an optimal one for the algorithm to seek a good balance between detection rate and processing time.
He, X.S., Wang, H., Jia, W., Wu, Q., Hur, N., Kim, J. & Hintz, T.B. 2007, 'Uniform Image Partitioning for Fractal Compression on Virtual Hexagonal Structure', International Journal of Information and Systems Science, vol. 3, no. 3, pp. 492-509.
View/Download from: UTSePress | Publisher's site
Hexagonal structure is different from the traditional square structure for image representation. The geometrical arrangement of pixels on hexag-onal structure can be described in terms of a hexagonal grid. Uniformly sepa-rating image into seven similar copies with a smaller scale has commonly been used for parallel and accurate image processing including image compression on hexagonal structure. However, all the existing hardware for capturing image and for displaying image are produced based on square architecture. It has become a serious problem affecting the advanced research based on hexagonal structure. Furthermore, the current techniques used for uniform separation of images on hexagonal structure do not coincide with the rectangular shape of images. This has been an obstacle in the use of hexagonal structure for image processing. In this paper, we briefly review a newly developed virtual hexagonal structure that is scalable. Based on this virtual structure, algorithms for uni-form image separation are presented. The virtual hexagonal structure retains image resolution during the process of image separation, and does not intro-duce distortion. Furthermore, images can be smoothly and easily transferred between the traditional square structure and the hexagonal structure while the image shape is kept in rectangle. As an application of image partitioning, we present a Fractal Image Compression (FIC) method on the virtual image struc- ture by adopting Fisher's basic FIC method on the traditional square image structure. The modifcation on the definition of range block and domain block is implemented in order to utilize the enhanced image structure. The results of the FIC approach applied to testing images are analyzed and show higher fidelity.
He, X.S., Jia, W., Wu, Q. & Hintz, T.B. 2006, 'Description of the cardiac movement using hexagonal image structures', Computerised medical imaging and graphics, vol. 30, no. 6-7, pp. 377-382.
View/Download from: UTSePress | Publisher's site
The most notable characteristic of the heart is its movement. Detection of dynamic information describing cardiac movement such as amplitude, speed and acceleration facilitates interpretation of normal and abnormal function. In recent years, the Omni-directional M-mode Echocardiography System (OMES) has been developed as a process that builds moving information from a sequence of echocardiography image frames. OMES detects cardiac movement through construction and analysis of Position+Time Grey Waveform (PTGW) images on some feature points of the boundaries of the ventricles. Image edge detection plays an important role in determining the feature boundary points and their moving directions as the basis for extraction of PTGW images+Spiral Architecture (SA) has proved efficient for image edge detection. SA is a hexagonal image structure in which an image is represented as a collection of hexagonal pixels. There are two operations called spiral addition and spiral multiplication defined on SA. They correspond to image translation and rotation, respectively. In this paper, we perform ventricle boundary detection based on SA using various defined chain codes. The gradient direction of each boundary point is determined at the same time. PTGW images at each boundary point are obtained through a series of spiral additions according to the directions of boundary points. Unlike the OMES system, our new approach is no longer affected by the translation movement of the heart. As its result, three curves representing the amplitude, speed and acceleration of cardiac movement can be easily drawn from the PTGW images obtained. Our approach is more efficient and accurate than OMES, and our results contain a more robust and complete description of cardiac motion.
Lin, Q., Zhang, L. & Jia, W. 2002, 'Omnidirectional Grey-Time Waveform System and Its Application in Ultrasound Echocardiography', Journal of Electronic Measurement and Instrument, vol. 16, no. 2, pp. 70-75.