Open Access. Powered by Scholars. Published by Universities.®
Physical Sciences and Mathematics Commons™
Open Access. Powered by Scholars. Published by Universities.®
- File Type
Articles 1 - 30 of 76
Full-Text Articles in Physical Sciences and Mathematics
Greedy Approximation Of Kernel Pca By Minimizing The Mapping Error, Peng Cheng, Wanqing Li, Philip Ogunbona
Greedy Approximation Of Kernel Pca By Minimizing The Mapping Error, Peng Cheng, Wanqing Li, Philip Ogunbona
Professor Philip Ogunbona
In this paper we propose a new kernel PCA (KPCA) speed-up algorithm that aims to find a reduced KPCA to approximate the kernel mapping. The algorithm works by greedily choosing a subset of the training samples that minimizes the mean square error of the kernel mapping between the original KPCA and the reduced KPCA. Experimental results have shown that the proposed algorithm is more efficient in computation and effective with lower mapping errors than previous algorithms.
Perceived Similarity And Visual Descriptions In Content-Based Image Retrieval, Yuan Zhong, Lei Ye, Wanqing Li, Philip Ogunbona
Perceived Similarity And Visual Descriptions In Content-Based Image Retrieval, Yuan Zhong, Lei Ye, Wanqing Li, Philip Ogunbona
Professor Philip Ogunbona
The use of low-level feature descriptors is pervasive in content-based image retrieval tasks and the answer to the question of how well these features describe users’ intention is inconclusive. In this paper we devise experiments to gauge the degree of alignment between the description of target images by humans and that implicitly provided by low-level image feature descriptors. Data was collected on how humans perceive similarity in images. Using images judged by humans to be similar, as ground truth, the performance of some MPEG-7 visual feature descriptors were evaluated. It is found that various descriptors play different roles in different …
Index-Compressed Vector Quantisation Based On Index Mapping, Jamshid Shanbehzadeh, Philip Ogunbona
Index-Compressed Vector Quantisation Based On Index Mapping, Jamshid Shanbehzadeh, Philip Ogunbona
Professor Philip Ogunbona
The authors introduce a novel coding technique which significantly improves the performance of the traditional vector quantisation (VQ) schemes at low bit rates. High interblock correlation in natural images results in a high probability that neighbouring image blocks are mapped to small subsets of the VQ codebook, which contains highly correlated codevectors. If, instead of the whole VQ codebook, a small subset is considered for the purpose of encoding neighbouring blocks, it is possible to improve the performance of traditional VQ schemes significantly. The performance improvement obtained with the new method is about 3dB on average when compared with traditional …
Emotional States Control For On-Line Game Avatars, Ce Zhan, Wanqing Li, Farzad Safaei, Philip Ogunbona
Emotional States Control For On-Line Game Avatars, Ce Zhan, Wanqing Li, Farzad Safaei, Philip Ogunbona
Professor Philip Ogunbona
Although detailed animation has already been achieved in a number of Multi-player On-line Games (MOGs), players have to use text commands to control emotional states of avatars. Some systems have been proposed to implement a real-time automatic system facial expression recognition of players. Such systems can then be used to control avatars emotional states by driving the MOG's "animation engine" instead of text commands. Some of the challenges of such systems is the ability to detect and recognize facial components from low spatial resolution face images. In this paper a system based on an improved face detection method of Viola …
Modelling Of Color Cross-Talk In Cmos Image Sensors, Wanqing Li, Philip Ogunbona, Yan Shi, Igor Kharitonenko
Modelling Of Color Cross-Talk In Cmos Image Sensors, Wanqing Li, Philip Ogunbona, Yan Shi, Igor Kharitonenko
Professor Philip Ogunbona
This paper presents a way to model the cross-talk effect in CMOS image sensors. Two algorithms are derived from the model; both of them work on the Bayer raw data and have low computational complexity. Experiments on Macbeth color chart and real images have shown the effectiveness of the modeling to eliminate the cross-talk effect and produce better quality images with traditional color interpolation and correction algorithms designed for CCD image sensors.
Cmos Sensor Cross-Talk Compensation For Digital Cameras, Wanqing Li, Philip Ogunbona, Yu Shi, Igor Kharitonenko
Cmos Sensor Cross-Talk Compensation For Digital Cameras, Wanqing Li, Philip Ogunbona, Yu Shi, Igor Kharitonenko
Professor Philip Ogunbona
This paper presents two algorithms for removing the cross-talk effect in CMOS sensor based color-imaging systems. The algorithms work on the Bayer raw data and have low computational complexity. Experimental results on Macbeth color chart and real images demonstrated that both algorithms can effectively eliminate the cross-talk effect and produce better quality images with conventional color interpolation and correction algorithms designed for CCD image sensors. Complexity of the algorithms is also analyzed.
Application Of Visual Modelling In Image Restoration And Colour Image Processing, Aziz Qureshi, Philip Ogunbona
Application Of Visual Modelling In Image Restoration And Colour Image Processing, Aziz Qureshi, Philip Ogunbona
Professor Philip Ogunbona
This paper describes the application of human visual models in (i) defining a visually uniform colour representation space and (ii) the formulation of visually weighted Kalman filtering for image restoration. The former being useful in colour image quantisation and compression. For (i), the uniformity of chromaticity differences at the ouptut of Frei ’s colour vision model [3] is tested and compensated for by using MacAdam’s uniform chromaticity space. For (ii), the dynamical image model of the Kalman filter is visually weighted using the frequency response of Stockham’s model [l] of human vision.
Channel-Optimized Vector Trellis Source Coding For The Awgn Channel, Philip Secker, Philip Ogunbona
Channel-Optimized Vector Trellis Source Coding For The Awgn Channel, Philip Secker, Philip Ogunbona
Professor Philip Ogunbona
A channel-optimised (joint source and channel) trellis source coder is designed for the AWGN channel. The optimum decoder is a non-linear function of the real channel information. The extension to 2D vector alphabets coupled with modifications to the signal space are found to improve performance. Favourable comparisons are made against a trellis source coder/TCM system.
Visual Perceptual Process Model And Object Segmentation, Wanqing Li, P. Ogunbona, Lei Ye, Igor Kharitonenko
Visual Perceptual Process Model And Object Segmentation, Wanqing Li, P. Ogunbona, Lei Ye, Igor Kharitonenko
Professor Philip Ogunbona
Modeling human visual process is crucial for automatic object segmentation that is able to produce consistent results to human perception. Based on the latest understanding of how human performs the task of extracting objects from images, we proposed a graph-based computational framework to model the visual process. The model supports the hierarchical nature of human visual perception and consists of the key steps of human visual perception including pre-attentive (pre-constancy) grouping, figure-and-ground organization, and attentive (post-constancy) grouping. A divide-and-conquer implementation of the model based on the concept of shortest spanning tree (SST) has demonstrated the potential of the model for …
Industrial Computer Vision Using Undefined Feature Extraction, Phil Evans, John A. Fulcher, Philip Ogunbona
Industrial Computer Vision Using Undefined Feature Extraction, Phil Evans, John A. Fulcher, Philip Ogunbona
Professor Philip Ogunbona
This paper presents an application of computer The implementation and operation of the system is vision in a real-world uncontrolled environment found at BHP Steel Port Kembla. The task is visual identification of torpedo ladles at a Blast Furnace wlahdilceh. is achieved by reading numbers attached to each 3. IMPLEMENTATION Number recognition is achieved through use of feature extraction using a Multi-Layer Perceptron (MLP) Artificial Neural Network (ANN). The novelty in the method used in this application is that the features the MLP is being trained to extract are undefined before the MLP is initialised. The results of the MLP …
Kernel Pca Of Hog Features For Posture Detection, Peng Cheng, Wanqing Li, Philip Ogunbona
Kernel Pca Of Hog Features For Posture Detection, Peng Cheng, Wanqing Li, Philip Ogunbona
Professor Philip Ogunbona
Motivated by the non-linear manifold learning ability of the Kernel Principal Component Analysis (KPCA), we propose in this paper a method for detecting human postures from single images by employing KPCA to learn the manifold span of a set of HOG features that can effectively represent the postures. The main contribution of this paper is to apply the KPCA as a non-linear learning and open-set classification tool, which implicitly learns a smooth manifold from noisy data that scatter over the feature space. For a new instance of HOG feature, its distance to the manifold that is measured by its reconstruction …
Index Factorised Image Adaptive Vector Quantisation, Jamshid Shanbehzadeh, Philip Ogunbona
Index Factorised Image Adaptive Vector Quantisation, Jamshid Shanbehzadeh, Philip Ogunbona
Professor Philip Ogunbona
No abstract provided.
On Multiple Watermarking, Nicholas Paul Sheppard, Reihaneh Safavi-Naini, Philip Ogunbona
On Multiple Watermarking, Nicholas Paul Sheppard, Reihaneh Safavi-Naini, Philip Ogunbona
Professor Philip Ogunbona
Mintzer and Braudaway once asked: If one watermark is good, are more better? In this paper, we discuss some techniques for embedding multiple watermarks into a single multimedia object and report some observations on implementations of these techniques.
Methods Of Channel-Optimised Trellis Source Coding For The Awgn Channel, Philip Secker, Philip Ogunbona
Methods Of Channel-Optimised Trellis Source Coding For The Awgn Channel, Philip Secker, Philip Ogunbona
Professor Philip Ogunbona
Improvements to channel-optimised trellis source coding for the AWGN channel are obtained by using, in various forms, real or ‘soft’ channel information. The proposed 1 bit/sample systems use a channel-optimised encoder matched to 1) a simple decision feedback detector, 2) an expanded codebook with 2-bit quantized information and 3) an optimum non-linear estimator decoder. The third system is further improved by considering vector alphabets and both constant and average energy constrained 2D signal constellations.
Face To Face Communications In Multiplayer Online Games: A Real-Time System, Ce Zhan, Wanqing Li, Farzad Safaei, Philip Ogunbona
Face To Face Communications In Multiplayer Online Games: A Real-Time System, Ce Zhan, Wanqing Li, Farzad Safaei, Philip Ogunbona
Professor Philip Ogunbona
Multiplayer online games (MOG) bring HCI into a new era of human-human interactions in computer world. Although current MOG provide more interactivity and social interaction in the virtual world, natural facial expression as a key factor in emulating face to face communications has been neglected by game designers. In this work, we propose a real-time automatic system to recognize players’ facial expressions, so that the recognition results can be used to drive the MOG’s “facial expression engine” instead of “text commands”. Our major contributions are the evaluation, improvement and efficient implementation of existing algorithms to build a real-time system that …
Human Detection Based On Weighted Template Matching, Duc Thanh Nguyen, Philip Ogunbona, Wanqing Li
Human Detection Based On Weighted Template Matching, Duc Thanh Nguyen, Philip Ogunbona, Wanqing Li
Professor Philip Ogunbona
This paper proposes a new two-stage human detection method involving matching and verification. A Bayesian framework is developed to verify the matching score obtained from a weighted distance measure. Performance evaluation indicates that the proposed method is able to utilize the flexible matching scheme and produce superior true positive, true negative and low misclassification rates.
Finding Distinctive Facial Areas For Face Recognition, Ce Zhan, Wanqing Li, Philip O. Ogunbona
Finding Distinctive Facial Areas For Face Recognition, Ce Zhan, Wanqing Li, Philip O. Ogunbona
Professor Philip Ogunbona
One of the key issues for local appearance based face recognition methods is that how to find the most discriminative facial areas. Most of the existing methods take the assumption that anatomical facial components, such as the eyes, nose, and mouth, are the most useful areas for recognition. Other more elaborate methods locate the most salient parts within the face according to a pre-specified criterion. In this paper, a novel method is proposed to identify the discriminative facial areas for face recognition. Unlike the existing methods that only analyze the given face, the proposed method identifies the distinctive areas of …
Compression Performance Of Jpeg Encryption Scheme, C. Kailasanathan, R. Safavi-Naini, P. Ogunbona
Compression Performance Of Jpeg Encryption Scheme, C. Kailasanathan, R. Safavi-Naini, P. Ogunbona
Professor Philip Ogunbona
Recent development in the Internet and Web based technologies require faster communication of multimedia data in a secure form. A number of encryption schemes for MPEG have been proposed. In this paper, we evaluate the compression performance of JPEG which has been encrypted with the zig-zag permutation algorithm, suggest a security enhancement to the scheme, and propose an alternative to entropy coding recommended by JPEG to compensate for the compression drop occurring due to permutation.
Human Detection With Contour-Based Local Motion Binary Patterns, Duc Thanh Nguyen, Philip Ogunbona, Wanqing Li
Human Detection With Contour-Based Local Motion Binary Patterns, Duc Thanh Nguyen, Philip Ogunbona, Wanqing Li
Professor Philip Ogunbona
This paper presents a human detection method using contour- based local motion features. The local motion is encoded using a variant of the popular Local Binary Pattern (LBP) called Non-Redundant Local Binary Pattern (NRLBP) descriptor computed on the difference image of two consecutive frames. In addition, the local motion features are extracted along the human's boundary contour. Localising features on the contours has the advantage of utilizing a precise human shape description. A motivation of the proposed method is that most of informative movements are performed on boundary contours of the body parts, e.g. legs of pedestrians. Evaluation of the …
High-Capacity Steganography Using A Shared Colour Palette, G. Brisbane, R. Safavi-Naini, P. Ogunbona
High-Capacity Steganography Using A Shared Colour Palette, G. Brisbane, R. Safavi-Naini, P. Ogunbona
Professor Philip Ogunbona
Seppanen, Makela and Keskinarkaus (SMK) have proposed a high-capacity steganographic technique to conceal information within a colour image. The technique is significant because of the high volume of data that is embedded into pixels but it results in a high level of noise and so the quality of the resulting image is not acceptable. A new type of coding structure is proposed, which maintains a high capacity but lowers the level of noise. Secondly, an adaptive algorithm is used to identify pixel values that have a high capacity to distortion ratio. Also the maximum size of the coding structures is …
Image Compression Based On Genealogical Relation Of The Tsvq Indices, Jamshid Shanbehzadeh, Philip Ogunbona, Abdoihosein Sarafzadeh
Image Compression Based On Genealogical Relation Of The Tsvq Indices, Jamshid Shanbehzadeh, Philip Ogunbona, Abdoihosein Sarafzadeh
Professor Philip Ogunbona
The indices obtained by tree-structured vector quantisation (TSVQ) have an interesting property that enables them to give information about the correlation between two image blocks. Iftwo image blocks are highly correlated, they may have an identical index, or the same ancestors. The existence of high inter-block correlation in natural images results in having neighboring blocks with the same genealogy. This characteristic can be used to compress the indices. This paper introduces a novel method to exploit the genealogical relation between the image block indices obtained from a TSVQ. The performance of this scheme in terms of PSNR versus average rate …
A Secure And Flexible Authentication System For Digital Images, Takeyuki Uehara, Reihaneh Safavi-Naini, Philip Ogunbona
A Secure And Flexible Authentication System For Digital Images, Takeyuki Uehara, Reihaneh Safavi-Naini, Philip Ogunbona
Professor Philip Ogunbona
Authentication of image data is a challenging task. Unlike data authentication systems that detect a single bit change in the data, image authentication systems must remain tolerant to changes resulting from acceptable image processing or compression algorithms while detecting malicious tampering with the image. Tolerance to the changes due to lossy compression systems is particularly important because in the majority of cases images are stored and transmitted in compressed form, and so it is important for verification to succeed if the compression is within the allowable range. In this paper we consider an image authentication system that generates an authentication …
On The Computational Complexity Of The Lbg And Pnn Algorithms, Jamshid Shanbehzadeh, Philip Ogunbona
On The Computational Complexity Of The Lbg And Pnn Algorithms, Jamshid Shanbehzadeh, Philip Ogunbona
Professor Philip Ogunbona
This correspondence compares the computational complexity of the pair-wise nearest neighbor (PNN) and Linde–Buzo–Gray (LBG) algorithms by deriving analytical expressions for their computational times. It is shown that for a practical codebook size and training vector sequence, the LBG algorithm is indeed more computationally efficient than the PNN algorithm.
A New Divide And Conquer Algorithm For Graph-Based Image And Video Segmentation, Wanqing Li, M. Shi, P. Ogunbona
A New Divide And Conquer Algorithm For Graph-Based Image And Video Segmentation, Wanqing Li, M. Shi, P. Ogunbona
Professor Philip Ogunbona
The concept of the Shortest (or Minimum) Spanning Tree (SST)and Recursive SST (RSST) of an undirected weighted graph has been successfully applied in image segmentation and edge detection. This paper presents a divide-and-conquer approach for (R)SST based image segmentation in order to over-come the problem of high computational complexity associated with conventional graph algorithms. In the simplest form, the proposed approach, block-based RSST (BRSST), first divides the image into rectangular blocks, finds the (R)SST of each block individually using conventional graph algorithms and, then, merges the (R) SSTs of all image blocks to form an (R)SST of the entire image. …
Secure Multimedia Authoring With Dishonest Collaborators, N. P. Sheppard, R. Safavi-Naini, P. Ogunbona
Secure Multimedia Authoring With Dishonest Collaborators, N. P. Sheppard, R. Safavi-Naini, P. Ogunbona
Professor Philip Ogunbona
Many systems have been proposed for protecting the intellectual property of multimedia authors and owners from the public at large, who have access to the multimedia only after it is published. In this paper, we consider the problem of protecting authors' intellectual property rights from insiders, such as collaborating authors and producers, who interact with the creative process before publication. We describe the weaknesses of standard proof-of-ownership watermarking approaches against dishonest insiders, and propose several possible architectures for systems that avoid these weaknesses. We further show how these architectures can be adapted for fingerprinting in the presence of dishonest insiders.
Performance Enhancement For Fuzzy Adaptive Resonance Theory (Art) Neural Networks, Golshah Naghdy, Jiazhao Wang, Philip Ogunbona
Performance Enhancement For Fuzzy Adaptive Resonance Theory (Art) Neural Networks, Golshah Naghdy, Jiazhao Wang, Philip Ogunbona
Professor Philip Ogunbona
A modified fuzzy adaptive resonance theory neural network (ART) is used as a classifier for a texture recognition system. The system consists of a wavelet based low level feature detector and a high level ART classifier. The performance improvement is measured in terms of identification accuracy and computational burden.
An Audio Representation For Content Based Retrieval, Kathy Melih, Ruben Gonzalez, Philip Ogunbona
An Audio Representation For Content Based Retrieval, Kathy Melih, Ruben Gonzalez, Philip Ogunbona
Professor Philip Ogunbona
Despite: the increasing interest in multimedia data retrieval audio data has received little attention. This is due, not to a lack of interest but rather to unique difficulties posed by the medium. In particular existing unstructured audio representations do not easily lend themselves to content based retrieval and especially browsing. This paper aims to address hs oversight by developing an audio representation that provides direct support for browsing and content based retrieval. This support is the result of a structured representation based on psychoacoustic ptincip1.e~in which salient attributes of audio are directly accessible. In addition, the representation is compact thus …
Index Compressed Tree-Structured Vector Quantisation, Jamshid Shanbehzadeh, Philip Ogunbona
Index Compressed Tree-Structured Vector Quantisation, Jamshid Shanbehzadeh, Philip Ogunbona
Professor Philip Ogunbona
This paper introduces a novel coding scheme based on Tree-Structured Vector Quantisation (TSVQ) scheme for image compression. The genealogical relationship among the indices of the neighbouring blocks generated from vector quantisation is exploited to improve the coding performance of TSVQ. The proposed coding scheme provides about 3.5 dB improvement over the basic TSVQ scheme and outperforms VQ schemes with memory and JPEG coding standard at low bit-rates. In addition its performance is comparable with address VQ but with much less complexity.
Facial Expression Recognition For Multiplayer Online Games, Ce Zhan, Wanqing Li, Philip O. Ogunbona, Farzad Safaei
Facial Expression Recognition For Multiplayer Online Games, Ce Zhan, Wanqing Li, Philip O. Ogunbona, Farzad Safaei
Professor Philip Ogunbona
The Multiplayer Online Game (MOG) becomes more popular than any other types of computer games for its collaboration, communication and interaction ability. However, compared with the ordinary human communication, the MOG still has many limitations, especially in communication using facial expressions. Although detailed facial animation has already been achieved in a number of MOGs, players have to use text commands to control avatars expressions. In this paper, we briefly review the state of the art in facial expression recognition and propose an automatic expression recognition system that can be integrated into a MOG to control the avatar’s facial expressions. We …
An Mpeg Tolerant Authentication System For Video Data, Takeyuki Uehara, R. Safavi-Naini, P. Ogunbona
An Mpeg Tolerant Authentication System For Video Data, Takeyuki Uehara, R. Safavi-Naini, P. Ogunbona
Professor Philip Ogunbona
We propose a secure video authentication algorithm that is tolerant to visual degradation due to MPEG lossy compression to a designed level. The authentication process generates a tag that is sent with video data and the level of protection can be adjusted so that longer tags are used for higher security, and that the protection is distributed such that higher security is provided for regions of interest in the image. The computation required for authentication and verification can be largely performed as part of MPEG compression and so generation and verification of the tag can be integrated into the compression …