In ordinary image coding, the same image quality is obtained in all parts of an image. If it is known that there is only one viewer, and where in the image that viewer is focusing, the quality can be degraded in other parts of the image without incurring any perceptible coding artefacts. This master's thesispresents a coding scheme where an image is segmented into homogeneous regions which are then separately coded, and where knowledge about the user's focus point is used to obtain further data reduction. It is concluded that the coding performance does not quite reach the levels attained when applying focus-based quality degradation to coding schemes not based on segmentation.

Given the recent abundance of brand choice data from scanner panels market researchers have neglected the measurement and analysis of perceptions. Heterogeneity of perceptions is still a largely unexplored issue in market structure and segmentation studies. Over the last decade various parametric approaches toward modelling segmented perception-preference structures such as combined MDS and Latent Class procedures have been introduced. These methods, however, are not taylored for qualitative data describing consumers' redundant and fuzzy perceptions of brand images. A completely different method is based on topology-sensitive vector quantization (VQ) for consumers-by-brands-by-attributes data. It maps the segment-specific perceptual structures into bubble-pie-bar charts with multiple brand positions demonstrating perceptual distinctiveness or similarity. Though the analysis proceeds without any distributional assumptions it allows for significance testing. The application of exploratory and inferential data processing steps to the same data base is statistically sound and particularly attractive for market structure analysts. A brief outline of the VQ method is followed by a sample study with travel market data which proved to be particularly troublesome for conventional processing tools. (author's abstract) / Series: Report Series SFB "Adaptive Information Systems and Modelling in Economics and Management Science"

Image segmentation is an important step in image processing, with many applications such as pattern recognition, object detection, and medical image analysis. It is a technique that separates objects of interests from the background in an image. Geometric active contour is a recent image segmentation method that overcomes previous problems with snakes. It is an attractive method for medical image segmentation as it is able to capture the object of interest in one continuous curve. The theory and implementation details of geometric active contours are discussed in this work. The robustness of the algorithm is tested through a series of tests, involving both synthetic images and medical images. Curve leaking past boundaries is a common problem in cases of non-ideal edges. Noise is also problematic for the advancement of the curve. Smoothing and parameters selection are discussed as ways to help solve these problems. This work also explores the incorporation of the multi-resolution method of Gaussian pyramids into the algorithm. Multi-resolution methods, used extensively in the areas of denoising and edge-selection, can help capture the spatial structure of an image. Results show that similar to the multi-resolution methods applied to parametric active contours, the multi-resolution can greatly increase the computation without sacrificing performance. In fact, results show that with successive smoothing and sub-sampling, performance often improves. Although smoothing and parameter adjustment help improve the performance of geometric active contours, the edge-based approach is still localized and the improvement is limited. Region-based approaches are recommended for further work on active contours.

This thesis describes a probabilistic approach to imagesegmentation and interpretation. The focus of the investigation is the development of a systematic way of combining color, brightness, texture and geometric features extracted from an image to arrive at a consistent interpretation for each pixel in the image. The contribution of this thesis is thus the presentation of a novel framework for the fusion of extracted image features producing a segmentation of an image into relevant regions. Further, a solution to the sub-pixel mixing problem is presented based on solving a probabilistic linear program. This work is specifically aimed at interpreting and digitizing multi-spectral aerial imagery of the Earth's surface. The features of interest for extraction are those of relevance to environmental management, monitoring and protection. The presented algorithms are suitable for use within a larger interpretive system. Some results are presented and contrasted with other techniques. The integration of these algorithms into a larger system is based firmly on a probabilistic methodology and the use of statistical decision theory to accomplish uncertain inference within the visual formalism of a graphical probability model.

The rapid proliferation of 3D volume data, including MRI and CT scans, is prompting the search within computer graphics for more effective volume visualisation techniques. Partially because of the traditional association with medical subjects, concepts borrowed from the domain of scientific illustration show great promise for enriching volume visualisation. This thesis describes the first general system dedicated to creating user-directed, variable-detail, scientific illustrations directly from volume data. In particular, using volume segmentation for explicit abstraction in non-photorealistic volume renderings is a new concept. The unique challenges and opportunities of volume data require rethinking many non-photorealistic algorithms that traditionally operate on polygonal meshes. The resulting 2D images are qualitatively different from but complementary to those normally seen in computer graphics, and inspire an analysis of the various artistic implications of volume models for scientific illustration.

The early detection of prostate cancer plays a significant role in the success of treatment and outcome. To detect prostate cancer, imaging modalities such as TransRectal UltraSound (TRUS) and Magnetic Resonance Imaging (MRI) are relied on. MRI images are more comprehensible than TRUS images which are corrupted by noise such as speckles and shadowing. However, MRI screening is costly, often unavailable in many community hospitals, time consuming, and requires more patient preparation time. Therefore, TRUS is more popular for screening and biopsy guidance for prostate cancer. For these reasons, TRUS images are chosen in this research. Radiologists first segment the prostate image from ultrasound image and then identify the hypoechoic regions which are more likely to exhibit cancer and should be considered for biopsy. In this thesis, the focus is on prostate segmentation and on Regions of Interest (ROI)segmentation. First, the extraneous tissues surrounding the prostate gland are eliminated. Consequently, the process of detecting the cancerous regions is focused on the prostate gland only. Thus, the diagnosing process is significantly shortened. Also, segmentation techniques such as thresholding, region growing, classification, clustering, Markov random field models, artificial neural networks (ANNs), atlas-guided, and deformable models are investigated. In this dissertation, the deformable model technique is selected because it is capable of segmenting difficult images such as ultrasound images. Deformable models are classified as either parametric or geometric deformable models. For the prostate segmentation, one of the parametric deformable models, Gradient Vector Flow (GVF) deformable contour, is adopted because it is capable of segmenting the prostate gland, even if the initial contour is not close to the prostate boundary. The manual segmentation of ultrasound images not only consumes much time and effort, but also leads to operator-dependent results. Therefore, a fully automatic prostate segmentation algorithm is proposed based on knowledge-based rules. The new algorithm results are evaluated with respect to their manual outlining by using distance-based and area-based metrics. Also, the novel technique is compared with two well-known semi-automatic algorithms to illustrate its superiority. With hypothesis testing, the proposed algorithm is statistically superior to the other two algorithms. The newly developed algorithm is operator-independent and capable of accurately segmenting a prostate gland with any shape and orientation from the ultrasound image. The focus of the second part of the research is to locate the regions which are more prone to cancer. Although the parametric dynamic contour technique can readily segment a single region, it is not conducive for segmenting multiple regions, as required in the regions of interest (ROI) segmentation part. Since the number of regions is not known beforehand, the problem is stated as 3D one by using level set approach to handle the topology changes such as splitting and merging the contours. For the proposed ROI segmentation algorithm, one of the geometric deformable models, active contours without edges, is used. This technique is capable of segmenting the regions with either weak edges, or even, no edges at all. The results of the proposed ROI segmentation algorithm are compared with those of the two experts' manual marking. The results are also compared with the common regions manually marked by both experts and with the total regions marked by either expert. The proposed ROI segmentation algorithm is also evaluated by using region-based and pixel-based strategies. The evaluation results indicate that the proposed algorithm produces similar results to those of the experts' manual markings, but with the added advantages of being fast and reliable. This novel algorithm also detects some regions that have been missed by one expert but confirmed by the other. In conclusion, the two newly devised algorithms can assist experts in segmenting the prostate image and detecting the suspicious abnormal regions that should be considered for biopsy. This leads to the reduction the number of biopsies, early detection of the diseased regions, proper management, and possible reduction of death related to prostate cancer.

Object tracking is an important problem with wide ranging applications. The purpose is to detect object contours and track their motion in a video. Issues of concern are to be able to map objects correctly between two frames, and to be able to track through occlusion. This thesis discusses a novel framework for the purpose of object tracking which is inspired from image registration and segmentation models. Occlusion of objects is also detected and handled in this framework in an appropriate manner. The main idea of our tracking framework is to reconstruct the sequence of images in the video. The process involves deforming all the objects in a given image frame, called the initial frame. Regularization terms are used to govern the deformation of the shape of the objects. We use elastic and viscous fluid model as the regularizer. The reconstructed frame is formed by combining the deformed objects with respect to the depth ordering. The correct reconstruction is selected by parameters that minimize the difference between the reconstruction and the consecutive frame, called the target frame. These parameters provide the required tracking information, such as the contour of the objects in the target frame including the occluded regions. The regularization term restricts the deformation of the object shape in the occluded region and thus gives an estimate of the object shape in this region. The other idea is to use a segmentation model as a measure in place of the frame difference measure. This is separate from image segmentation procedure, since we use the segmentation model in a tracking framework to capture object deformation. Numerical examples are presented to demonstrate tracking in simple and complex scenes, alongwith occlusion handling capability of our model. Segmentation measure is shown to be more robust with regard to accumulation of tracking error.

The accurate detection of static object boundaries such as contours or surfaces and dynamic tunnels of moving objects via deformable models is an ongoing research topic in computer vision. Most deformable models attempt to converge towards a desired solution by minimizing the sum of internal (prior) and external (measurement) energy terms. Such an approach is elegant, but frequently mis-converges in the presence of noise or complex boundaries and typically requires careful semi-dependent parameter tuning and initialization. Furthermore, current deformable model based approaches are computationally demanding which precludes real-time use. To address these limitations, a decoupled deformable model (DDM) is developed which optimizes the two energy terms separately. Essentially, the DDM consists of a measurement update step, employing a Hidden Markov Model (HMM) and Maximum Likelihood (ML) estimator, followed by a separate prior step, which modifies the updated deformable model based on the relative strengths of the measurement uncertainty and the non-stationary prior. The non-stationary prior is generated by using a curvature guided importance sampling method to capture high curvature regions. By separating the measurement and prior steps, the algorithm is less likely to mis-converge; furthermore, the use of a non-iterative ML estimator allows the method to converge more rapidly than energy-based iterative solvers. The full functionality of the DDM is developed in three phases. First, a DDM in 2D called the decoupled active contour (DAC) is developed to accurately identify the boundary of a 2D object in the presence of noise and background clutter. To carry out this task, the DAC employs the Viterbi algorithm as a truncated ML estimator, curvature guided importance sampling as a non-stationary prior generator, and a linear Bayesian estimator to fuse the non-stationary prior with the measurements. Experimental results clearly demonstrate that the DAC is robust to noise, can capture regions of very high curvature, and exhibits limited dependence on contour initialization or parameter settings. Compared to three other published methods and across many images, the DAC is found to be faster and to offer consistently accurate boundary identification. Second, a fast decoupled active contour (FDAC) is proposed to accelerate the convergence rate and the scalability of the DAC without sacrificing the accuracy by employing computationally efficient and scalable techniques to solve the three primary steps of DAC. The computational advantage of the FDAC is demonstrated both experimentally and analytically compared to three computationally efficient methods using illustrative examples. Finally, an extension of the FDAC from 2D to 3D called a decoupled active surface (DAS) is developed to precisely identify the surface of a volumetric 3D image and the tunnel of a moving 2D object. To achieve the objectives of the DAS, the concepts of the FDAC are extended to 3D by using a specialized 3D deformable model representation scheme and a computationally and storage efficient estimation scheme. The performance of the DAS is demonstrated using several natural and synthetic volumetric images and a sequence of moving objects.

Conventional ultrasonographic examination of the bovine ovary is based on a sequence of two-dimensional (2D) cross-section images. Day-to-day estimation of the number, size, shape and position of the ovarian follicles is one of the most important aspects of ovarian research. Computer-assisted follicle segmentation of ovarian volume can relieve physicians from the tedious manual detection of follicles, provide objective assessment of spatial relationships between the ovarian structures and therefore has the potential to improve accuracy. Modern segmentation procedures are performed on 2D images and the three-dimensional (3D) visualization of follicles is obtained from the reconstruction of a sequence of 2D segmented follicles. <p>The objective of this study was to develop a semi-automatic 3D follicle segmentation method based on seeded region growing. The 3D datasets were acquired from a sequence of 2D ultrasound images and the ovarian structures were segmented from the reconstructed ovarian volume in a single step. A seed is placed manually in each follicle and the growth of the seed is controlled by the algorithm using a combination of average grey-level, standard deviation of the intensity, newly-developed volumetric comparison test and a termination criterion. One important contribution of this algorithm is that it overcomes the boundary leakage problem of follicles of conventional 2D segmentation procedures. The results were validated against the aspiration volume of follicles, the manually detected follicles by an expert and an existing algorithm.<p>We anticipate that this algorithm will enhance follicular assessment based on current ultrasound techniques in cases when large numbers of follicles (e.g. ovarian superstimulation) obviate accurate counting and size measurement.

The objective of this thesis is to detect and identify moving objects in a video sequence. The currently available techniques for motion estimation can be broadly categorized into two main classes: block matching methods and optical flow methods.<p>This thesis investigates the different motion estimation algorithms used for video processing applications. Among the available motion estimation methods, the Lucas Kanade Optical Flow Algorithm has been used in this thesis for detection of moving objects in a video sequence. Derivatives of image brightness with respect to x-direction, y-direction and time t are calculated to solve the Optical Flow Constraint Equation. The algorithm produces results in the form of horizontal and vertical components of optical flow velocity, u and v respectively. This optical flow velocity is measured in the form of vectors and has been used to segment the moving objects from the video sequence. The algorithm has been applied to different sets of synthetic and real video sequences.<p>This method has been modified to include parameters such as neighborhood size and Gaussian pyramid filtering which improve the motion estimation process. The concept of Gaussian pyramids has been used to simplify the complex video sequences and the optical flow algorithm has been applied to different levels of pyramids. The estimated motion derived from the difference in the optical flow vectors for moving objects and stationary background has been used to segment the moving objects in the video sequences. A combination of erosion and dilation techniques is then used to improve the quality of already segmented content.<p>The Lucas Kanade Optical Flow Algorithm along with other considered parameters produces encouraging motion estimation and segmentation results. The consistency of the algorithm has been tested by the usage of different types of motion and video sequences. Other contributions of this thesis also include a comparative analysis of the optical flow algorithm with other existing motion estimation and segmentation techniques. The comparison shows that there is need to achieve a balance between accuracy and computational speed for the implementation of any motion estimation algorithm in real time for video surveillance.