Search

For high-dimensional classification tasks such as face recognition, the number of samples is smaller than the dimensionality of the samples. In such cases, a problem encountered in Linear Discriminant Analysis-based (LDA) methods for dimension reduction is what is known as the Small Sample Size (SSS) problem. Recently, a number of approaches that attempt to solve the SSS problem have been proposed in the literature. In this paper, a different way of solving the SSS problem compared to these is proposed. It is one that employs a dissimilarity representation method where an object is represented based on the dissimilarity measures among representatives extracted from training samples instead of from the feature vector itself. Thus, by appropriately selecting representatives and by defining the dissimilarity measure, it is possible to reduce the dimensionality and achieve a better classification performance in terms of both speed and accuracy. Apart from utilizing the dissimilarity representation, in this paper simultaneously employing a fusion technique is also proposed in order to increase the classification accuracy. The rationale for this is explained in the paper. The proposed scheme is completely different from the conventional ones in terms of the computation of the transformation matrix, as well as in controlling the number of dimensions. The present experimental results, which to the best of the authors' knowledge, are the first such reported results, demonstrate that the proposed mechanism achieves nearly identical efficiency results in terms of the classification accuracy compared with the conventional LDA-extension approaches for well-known face databases involving AT&T and Yale databases. [ABSTRACT FROM AUTHOR]

This paper proposes a method for detecting frontal pedestrian crossings and estimating its length from image data obtained with a single camera as a travel aid for the blind. It is important for the blind to know whether or not a frontal area is a crossing. The existence of a crossing is detected in two steps. In the first step, feature points for a crossing is extracted using Fisher criterion. In the second step, the existence of a crossing is detected by checking on the peoriodicity of white stripes on the road using projective invariants. Next, we propose a method for estimaing crossing length using extracted feature points. From the experimental results for evaluation, it is found that the existence of a crossing is successfully detected for all 173 real images which include 100 images with crossings and 73 images without crossing. The rms error of crossing length estimation for the 100 images is found 2.29m. [ABSTRACT FROM AUTHOR]

In most content-based image retrieval systems, the low level visual features such as color, texture and region play an important role. Variety of dissimilarity measures were introduced for an uniform quantization of visual features, or a histogram. However, a cluster-based representation, or a signature, has proven to be more compact and theoretically sound for the accuracy and robustness than a histogram. Despite of these advantages, so far, only a few dissimilarity measures have been proposed. In this paper, we present a novel dissimilarity measure for a random signature, Perceptually Modified Hausdorff Distance (PMHD), based on Hausdorff distance. In order to demonstrate the performance of the PMHD, we retrieve relevant images for some queries on real image database by using only color information. The precision vs. recall results show that the proposed dissimilarity measure generally outperforms all other dissimilarity measures on an unmodified commercial image database. [ABSTRACT FROM AUTHOR]

The cropping of image is one of most popular functions in current image editing software for general digital camera users. And image-based identifier system is needed for wide distribution of digital image products. In this paper, we propose new concept of visual identifier for digital photos and visual identifier system structure robust against especially cropped natural photos. Visual identifier is new concept with different ground truth rather than retrieval. In the proposed identifier system, local corner is used as main co-location position and local gradient histogram is applied to describe feature in each position. And for robust matching we use simple random sample consensus method. Since image cropping can be considered as a kind of translation, linear model is sufficient as geometric transform model. For experiment we make 11 kinds of modifications of original images and evaluate performance of the proposed algorithm. From experiment results, our proposed algorithm shows better performance relative to previous MPEG-7 visual descriptors. [ABSTRACT FROM AUTHOR]

This paper introduces a novel Gabor-based supervised locality preserving projection (GSLPP) method for face recognition. Locality preserving projection (LPP) is a recently proposed method for unsupervised linear dimensionality reduction. LPP seeks to preserve the local structure which is usually more significant than the global structure preserved by principal component analysis (PCA) and linear discriminant analysis (LDA). In this paper, we investigate its extension, called supervised locality preserving projection (SLPP), using class labels of data points to enhance its discriminant power in their mapping into a low dimensional space. The GSLPP method, which is robust to variations of illumination and facial expression, applies the SLPP to an augmented Gabor feature vector derived from the Gabor wavelet representation of face images. We performed comparative experiments of various face recognition schemes, including the proposed GSLPP method, principal component analysis (PCA) method, linear discriminant analysis (LDA) method, locality preserving projection method, the combination of Gabor and PCA method (GPCA) and the combination of Gabor and LDA method (GLDA). Experimental results on AR database and CMU PIE database show superior of the novel GSLPP method. [ABSTRACT FROM AUTHOR]

In this paper, we present AddCanny, an Anisotropic Diffusion and Dynamic reformulation of the Canny edge detector. The proposal provides two modifications to classical Canny detector. The first one consists of using an anisotropic diffusion filter instead of a Gaussian filter as Canny does in order to obtain better edge detection and location. The second one is the replacement of the hysteresis step by a dynamic threshold process, in order to reduce blinking effect of edges during successive frames and, therefore, generate more stable edges in sequences. Also, a new performance measurement based on the Euclidean Distance Transform to evaluate the consistency of computed edges is proposed. The paper includes experimental evaluations with different video streams that illustrate the advantages of AddCanny compared to classical Canny edge detector. [ABSTRACT FROM AUTHOR]

In many application scenarios, the use of Regions of Interest (ROIs) within video sequences is a useful concept. It is shown in this paper how Flexible Macroblock Ordering (FMO), defined in H.264/AVC as an error resilience tool, can be used for the coding arbitrary-shaped ROIs. In order to exploit the coding of ROIs in an H.264/AVC bitstream, a description-driven content adaptation framework is introduced that is able to extract the ROIs of a given bitstream. The results of a series of tests indicate that the ROI extraction process significantly reduces the bit rate of the bitstreams and increases the decoding speed. In case of a fixed camera and a static background, the impact of this reduction on the visual quality of the video sequence is negligible. Regarding the adaptation framework itself, it is shown that in all cases, the framework operates in real time and that it is suited for streaming scenarios by design. [ABSTRACT FROM AUTHOR]

In this paper we present our region merging algorithm that is built with special attention on speed but still includes all the necessary functionality to use a wide range of both region based and border based dissimilarity criteria. The algorithm includes a novel method to dynamically link the common borders between two segments during the region merging. The main part of the paper will concentrate on the efficient data structures and functions that are needed to obtain a fast region merging algorithm. Also, all the special situations that can occur in the segment topology are completely covered. We give a detailed report on the execution times of the algorithm and show some of the segmentation results we obtained. [ABSTRACT FROM AUTHOR]

Video classification provides an efficient way to manage and utilize the video data. Existing works on this topic fall into this category: enlarging the feature set until the classification is reliable enough. However, some features may be redundant or irrelevant. In this paper, we address the problem of choosing efficient feature set in video genre classification to achieve acceptable classification results but relieve computation burden significantly. A rough set approach is proposed. In comparison with existing works and the decision tree method, experimental results verify the efficiency of the proposed approach. [ABSTRACT FROM AUTHOR]

This paper presents a Road Detection and Classification algorithm for Driver Assistance Systems (DAS), which tracks several road lanes and identifies the type of lane boundaries. The algorithm uses an edge filter to extract the longitudinal road markings to which a straight lane model is fitted. Next, the type of right and left lane boundaries (continuous, broken or merge line) is identified using a Fourier analysis. Adjacent lanes are searched when broken or merge lines are detected. Although the knowledge of the line type is essential for a robust DAS, it has been seldom considered in previous works. This knowledge helps to guide the search for other lanes, and it is the basis to identify the type of road (one-way, two-way or freeway), as well as to tell the difference between allowed and forbidden maneuvers, such as crossing a continuous line. [ABSTRACT FROM AUTHOR]

The vision-based scene understanding technique that infers scene-interpreting contexts from real-world vision data has to not only deal with various uncertain environments but also reflect user's requests. Especially, learnability is a hot issue for the system. In this paper, we adopt a probabilistic approach to overcome the uncertainty, and propose an interactive learning method using combination of Bayesian network and logic network to reflect user's requirements in real-time. The logic network works for supporting logical inference of Bayesian network. In the result of some learning experiments using interactive data, we have confirmed that the proposed interactive learning method is useful for scene context reasoning. [ABSTRACT FROM AUTHOR]

This paper presents a problem of 3D images decomposition into spheres. The presented method is based on a fast Hough transform with an input image space division. An essential element of this method is the use of a clustering technique for partial data sets. The method simplifies the application of Hough transform to segmentation tasks as well as accelerates calculations considerably. [ABSTRACT FROM AUTHOR]

In adjusting for discrepancies between adjacent airborne laser scanning (ALS) data strips, previous studies generally used conjugate features such as points, lines, and surface objects; however, irrespective of the types of features employed, the adjustment process relies upon the existence of suitable conjugate features within the overlapping area and the ability of the employed method to detect and extract the features. These limitations make the process complex and sometimes limit the applicability of developed methodologies because of a lack of suitable features in overlapping areas. To address these problems, this paper presents a methodology that uses the topological characteristics of the terrain itself, which is represented by a contour tree (CT). This approach provides a robust methodology without the restrictions involved in methods that employ conjugate features. Our method also makes the overall process of adjustment generally applicable and automated. [ABSTRACT FROM AUTHOR]

The visual tracking of image regions is a research area of great interest within the computer vision community. One issue which has received quite attention in the last years has been the analysis of tracking algorithms which could be able to cope with changes in the appearance of the target region. Probably one of the most studied techniques proposed to model this appearance variability is that based on linear subspace models. Recently, efficient algorithms for fitting these models have been developed too, in many cases as an evolution of well studied approaches for the tracking of fixed appearance images. Additionally, new methods based on second order optimizers have been proposed for the tracking of targets with no appearance changes. In this paper we study the application of such techniques in the design of tracking algorithms for linear appearance models and compare their performance with three previous approaches. The achieved results show the efficiency of the use of second-order minimization in terms of both number of iterations required for convergence and convergence frequency. [ABSTRACT FROM AUTHOR]

This paper proposes an object-based, highly scalable, lossy-to-lossless coding approach for magnetic resonance (MR) images. The proposed approach, called OBHS-SPIHT, is based on the well known set partitioning in hierarchical trees (SPIHT) algorithm and supports both quality and resolution scalability. It progressively encodes each slice of the MR data set separately in a multiresolution fashion from low resolution to full resolution and in each resolution from low quality to lossless quality. To achieve more compression efficiency, the algorithm only encodes the main object of interest in the input data set, and ignores the unnecessary background. The experimental results show the efficiency of the proposed algorithm for multiresolution lossy-to-lossless MRI data coding. OBHS-SPIHT, is a very attractive coding approach for medical image information archiving and transmission applications especially over heterogeneous networks. [ABSTRACT FROM AUTHOR]

Skin detection consists in detecting human skin pixels from an image. In this paper we propose a new skin detection algorithm based on approximation of an image patch joint distribution, called Best-Tree distribution. A tree distribution model is more general than a bayesian network one. It can represent a joint distribution in an intuitive and efficient way. We assess the performance of our method on the Compaq database by measuring the Receiver Operating Characteristic curve and its under area. These measures have proved better performances of our model than the baseline one. [ABSTRACT FROM AUTHOR]

In this paper we propose a method that analyzes attack patterns and extracts watermark after restoring the watermarked image from the geometric attacks. The proposed algorithm consists of a spatial-domain key insertion part for attack analysis and a frequency-domain watermark insertion part using discrete wavelet transform. With the spatial-domain key extracted from the damaged image, the proposed algorithm analyzes distortion and finds the attack pattern. After restoring the damaged image, the algorithm extracts the embedded watermark. By using both spatial domain key and frequency domain watermark, the proposed algorithm can achieve robust watermark extraction against geometrical attacks and image compressions such as JPEG. [ABSTRACT FROM AUTHOR]

In this paper, we present a new scheme called Common Image method for face recognition. Our method has a couple of advantages over the conventional face recognition algorithms; one is that it can deal with the Small Sample Size(SSS) problem in LDA, and the other one is that it can achieve a better performance than traditional PCA by seeking the optimal projection vectors from image covariance matrix in a recognition task. As opposed to traditional PCA-based methods and LDA-based methods which employ Euclidean distance, Common Image methods adopted Assemble Matrix Distance(AMD) and IMage Euclidean Distance(IMED), by which the overall recognition rate could be improved. To test the recognition performance, a series of experiments were performed on CMU PIE, YaleB, and FERET face databases. The test results with these databases show that our Common Image method performs better than Discriminative Common Vector and 2DPCA-based methods. [ABSTRACT FROM AUTHOR]

In this paper we propose the use of Discrete Choice Analysis (DCA) for static facial expression classification. Facial expressions are described with expression descriptive units (EDU), consisting in a set of high level features derived from an active appearance model (AAM). The discrete choice model (DCM) is built considering the 6 universal facial expressions plus the neutral one as the set of the available alternatives. Each alternative is described by an utility function, defined as the sum of a linear combination of EDUs and a random term capturing the uncertainty. The utilities provide a measure of likelihood for a combinations of EDUs to represent a certain facial expression. They represent a natural way for the modeler to formalize her prior knowledge on the process. The model parameters are learned through maximum likelihood estimation and classification is performed assigning each test sample to the alternative showing the maximum utility. We compare the performance of the DCM classifier against Linear Discriminant Analysis (LDA), Generalized Discriminant Analysis (GDA), Relevant Component Analysis (RCA) and Support Vector Machine (SVM). Quantitative preliminary results are reported, showing good and encouraging performance of the DCM approach both in terms of recognition rate and discriminatory power. [ABSTRACT FROM AUTHOR]

In this paper, we propose an ownership identification scheme for digital images with binary and gray-level ownership statements. The proposed method uses the theories and properties of sampling distribution of means to satisfy the requirements of robustness and security. Essentially, our method will not really insert the ownership statement into the host image. Instead, the ownership share will be generated by the sampling method as a key to reveal the ownership statement. Besides, our method allows ownership statements to be of any size and avoids the hidden ownership statement to be destroyed by the latter ones. When the rightful ownership of the image needs to be identified, our method can reveal the ownership statement without resorting to the original image. Finally, several common attacks to the image will be held to verify the robustness and the security is also analyzed. [ABSTRACT FROM AUTHOR]

There have been many progresses in the stereo matching problem. However, some remaining problems still make stereo matching difficult. Occlusion is one of such problems. In this paper, we propose a new stereo matching model that addresses this problem by using an effective visibility constraint. By considering two images simultaneously, complex geometric configurations regarding the visibility of a pixel becomes simplified, so that the visibility constraint can be modeled as a pairwise MRF. Also since the proposed model enforces the consistency between two disparity maps, the final results become consistent with each other. Belief propagation is employed for the solution of the modeled pairwise MRF. Experimental results on the standard data set demonstrate the effectiveness of our approach. [ABSTRACT FROM AUTHOR]

Many industrial applications involve rotations. Different from traditional measurements, we propose a novel vision method based on image motion blur to estimate angular velocity and angular displacement in this paper. First, we transform 2D rotation to 1D translational motion by sectoring rotation blurred image. Then we use mathematical models in spatial and frequency domain to analyze translation blurred images. According to mathematical models in frequency domain, there is a series of dark parallel lines on the spectrum of the translation blurred image. These dark lines are exactly related to the velocity of translation and exposure time. Furthermore, based on the geometric relationship between rotation and translation, these dark lines are also related to angular velocity. Both simulation results and real experimental results, based on the proposed method in this paper, are provided. These results demonstrate the feasibility and efficiency of proposed method. [ABSTRACT FROM AUTHOR]

We propose a scanline energy minimization algorithm for stereo vision. The proposed algorithm differs from conventional energy minimization techniques in that it focuses on the relationship between local match cost solution and the energy minimization solution. The local solution is transformed into energy minimization solution through the optimization of the disparity discontinuity. In this paper, disparity discontinuities are targeted during the energy minimization instead of the disparities themselves. By eliminating and relocating the disparity discontinuities, the energy can be minimized in iterations of O(n) where n is the number of pixels. Although dynamic programming has been adequate for both speed and performance in the scan-line stereo, the proposed algorithm was shown to have better performance with comparable speed. [ABSTRACT FROM AUTHOR]

Choosing the adequate registration and simulation parameters in non-parametric image registration methods is an open question. There is no agreement about which are the optimal values (if any) for these parameters, since they depend on the images to be registered. As a result, in the literature the parameters involved in the registration process are arbitrarily fixed by the authors. The present paper is intended to address this issue. A two-step method is proposed to obtain the optimal values of these parameters, in terms of achieving in a minimum number of iterations the best trade-off between similarity of the images and smoothness of the transformation. These optimal values minimize the joint energy functional defined in a variational framework. We focus on the specific formulation of diffusion and curvature registration, but the exposed methodology can be directly applied to other non-parametric registration schemes. The proposed method is validated over different registration scenarios. [ABSTRACT FROM AUTHOR]

In the context of Universal Multimedia Access, efficient techniques are needed for the adaptation of video content. An important example is the reduction of the bitrate in order to satisfy the bandwidth constraints imposed by the network or the decoding capability of the terminal devices. Requantization transcoding is a fast technique for bitrate reduction, and has been successfully applied in previous video coding standards such as MPEG-2. In this paper, we examine requantization in H.264/AVC, focusing on the intra 16×16 prediction modes. Due to the newly introduced coding tools in H.264/AVC, new techniques are needed that are able to lower the bitrate at a minimal quality loss. We propose two novel architectures, one in the pixel domain and one in the frequency domain, that reuse the information from the incoming bitstream in an efficient way, and perform approximately equally well as a cascade of decoder and encoder. Due to their low computational complexity, the introduced architectures are highly suitable for on-the-fly video adaptation scenarios. [ABSTRACT FROM AUTHOR]

Content analysis of video and still images is attractive for multiple reasons, such as enabling content-based actions and image manipulation. This paper presents a new algorithm and feature model for blue-sky detection. The algorithm classifies the sky areas by computing a pixel-accurate sky probability. Such a probabilistic measure matches well with the requirements of typical video enhancement functions in TVs. This algorithm enables not only content-adaptive picture quality improvement, but also more advanced applications such as content-based annotation of, and retrieval from image and video databases. When compared to existing algorithms, our proposal shows considerable improvements in correct detection/rejection rate of sky areas, and an improved consistency of the segmentation results. [ABSTRACT FROM AUTHOR]

In this paper, we present a real-time sports analysis system, which not only recognizes the semantic events, but also concludes the behavior, like player's tactics. To this end, we propose an advanced multiple-player tracking algorithm, which addresses two improvements on practical problems: (1) updating of the player template so that it remains a good model over time, and (2) adaptive scaling of the template size depending on the player motion. In this algorithm, we obtain the initial locations of players in the first frame. The tracking is performed by considering both the kinematic constraints of the player and the color distribution of appearance, thereby achieving promising results. We demonstrate the performance of the proposed system by evaluating it for double tennis matches where the player count and the resulting occlusions are challenging. [ABSTRACT FROM AUTHOR]

Facial expression hallucination is an important approach to facial expression synthesis. Existing works mainly focused on synthesizing a static facial expression image given one face image with neutral expression. In this paper, we propose a novel two-level hierarchical fusion approach to hallucinate dynamic expression video sequences when given only one neutral expression face image. By fusion of local linear and global nonlinear subspace learning, the two-level approach provides a sound solution to organizing the complex video sample space. Experiments show that our approach generates reasonable facial expression sequences both in temporal domain and spatial domain with less artifact compared with existing works. [ABSTRACT FROM AUTHOR]

Because of limitation of low power and computational ability in mobile video devices, it is significant to develop energy-efficient wireless video compression methods for mobile video systems. In this paper, a new approach to low complexity wireless video compression is proposed, based on Wyner-Ziv coding theorem. The proposed method encodes video only by intracoding techniques and detection of regions of interest, without the complicated motion estimation and compensation in the mobile video terminal. Thus, the computational burden is obviously reduced and the requirement of low power can be satisfied in mobile video devices. Experimental results show that the proposed method is quite effective. [ABSTRACT FROM AUTHOR]

Employing block-based motion models in scalable video coding based on spatial-domain motion-compensated temporal filtering (SDMCTF) introduces artificial block-boundary discontinuities that adversely affect the video compression performance, particularly at low bit-rates. This paper focuses on the problem of deblocking in the context of SDMCTF-based video coding. One possible solution to this problem is the use of overlapped-block motion compensation (OBMC). An alternative solution is applying an adaptive deblocking filter, similar to the one used in H.264. With this respect, a novel adaptive deblocking filter, tailored to SDMCTF video coding is proposed. In terms of visual-quality, both approaches yield similar performance. However, adaptive deblocking is less complex than OBMC, as it requires up to 34% less processing time. Experimental results show that the two techniques significantly improve the subjective and objective quality of the decoded sequences, confirming the expected benefits brought by deblocking in SDMCTF-based video coding. [ABSTRACT FROM AUTHOR]

Intra coding in current hybrid video coding method has very important functionality like low delay in decoder, random access and error resilience. Unfortunately coding efficiency of intra frame is very low relative to inter frame coding because of mismatch between current block and its predicted block. In this paper, new intra luma prediction algorithm which improves intra coding efficiency is proposed. The proposed additional intra luma prediction mode uses collocated chroma pixels and weight values to estimate correct spatial pattern of coded block. From neighboring blocks, weight value between chroma and luma values is calculated and then the predicted luma block is obtained by multiplying calculated weight value and collocated upsampled chroma block. The proposed method is effective for complex or non-directional macroblocks and experimental results show that the efficiency of intra coding is increased up to 0.6 dB. [ABSTRACT FROM AUTHOR]

In this paper we adapt different techniques for image deconvolution, to the actual restoration of works of arts (mainly paintings and sculptures) from the baroque period. We use the special characteristics of these works in order to both restrict the strategies and benefit from those properties. We propose an algorithm which presents good results in the pieces we have worked. Due to the diversity of the period and the amount of artists who made it possible, the algorithms are too general even in this context. This is a first approach to the problem, in which we have assumed very common and shared features for the works of art. The flexibility of the algorithm, and the freedom to choose some parameters make it possible to adapt the problem to the knowledge that restorators in charge may have about a particular work. [ABSTRACT FROM AUTHOR]

This paper focuses on fuzzy image denoising techniques. In particular, we investigate the usage of fuzzy set theory in the domain of image enhancement using wavelet thresholding. We propose a simple but efficient new fuzzy wavelet shrinkage method, which can be seen as a fuzzy variant of a recently published probabilistic shrinkage method [1] for reducing adaptive Gaussian noise from digital greyscale images. Experimental results show that the proposed method can efficiently and rapidly remove additive Gaussian noise from digital greyscale images. Numerical and visual observations show that the performance of the proposed method outperforms current fuzzy non-wavelet methods and is comparable with some recent but more complex wavelets methods. We also illustrate the main differences between this version and the probabilistic version and show the main improvements in comparison to it. [ABSTRACT FROM AUTHOR]

Independent Component Analysis (ICA) has shown success in blind source separation. Its applications to remotely sensed images have been investigated recently. In this approach, a Linear Spectral Mixture (LSM) model is used to characterize spectral data. This model and the associated linear unmixing algorithms are based on the assumption that the spectrum for a given pixel in an image is a linear combination of the end-member spectra. The assumption that the abundances are mutually statistically independent random sources requires the separating matrix to be unitary. This paper considers a new approach, the Non Orthogonal Component Analysis (NOCA), which enables to relax this assumption. The experimental results demonstrate that the proposed NOCA provides a more effective technique for anomaly detection in hyperspectral imagery than the ICA approach. In particular, we highlight the fact that the difference between the performances of the two approaches increases when the number of bands decreases. [ABSTRACT FROM AUTHOR]

The nearest neighbor (NN) classifier is well suited for generic object recognition. However, it requires storing the complete training data, and classification time is linear in the amount of data. There are several approaches to improve runtime and/or memory requirements of nearest neighbor methods: Thinning methods select and store only part of the training data for the classifier. Efficient query structures reduce query times. In this paper, we present an experimental comparison and analysis of such methods using the ETH-80 database. We evaluate the following algorithms. Thinning: condensed nearest neighbor, reduced nearest neighbor, Baram's algorithm, the Baram-RNN hybrid algorithm, Gabriel and GSASH thinning. Query structures: kd-tree and approximate nearest neighbor. For the first four thinning algorithms, we also present an extension to k-NN which allows tuning the trade-off between data reduction and classifier degradation. The experiments show that most of the above methods are well suited for generic object recognition. [ABSTRACT FROM AUTHOR]

A fast method for the recognition and classification of informational traffic signs is presented in this paper. The aim is to provide an efficient framework which could be easily used in inventory and guidance systems. The process consists of several steps which include image segmentation, sign detection and reorientation, and finally traffic sign recognition. In a first stage, a static HSI colour segmentation is performed so that possible traffic signs can be easily isolated from the rest of the scene; secondly, shape classification is carried out so as to detect square blobs from the segmented image; next, each object is reoriented through the use of a homography transformation matrix and its potential axial deformation is corrected. Finally a recursive adaptive segmentation and a SVM-based recognition framework allow us to extract each possible pictogram, icon or symbol and classify the type of the traffic sign via a voting-scheme. [ABSTRACT FROM AUTHOR]

This paper presents a novel on-line learning procedure to be used in biologically realistic networks of integrate-and-fire neurons. The on-line adaptation is based on synaptic plasticity and changes in the network structure. Event driven computation optimizes processing speed in order to simulate networks with large number of neurons. The learning method is demonstrated on a visual recognition task and can be expanded to other data types. Preliminary experiments on face image data show the same performance as the optimized off-line method and promising generalization properties. [ABSTRACT FROM AUTHOR]

In this paper, we present a method for producing omnidirectional image signatures that are purposed to localize a mobile robot in an office environment. To solve the problem of perceptual aliasing common to the image based recognition approaches, we choose to build signatures that greatly vary between rooms and slowly vary inside a given room. To do so, an invariant approach has been developed, based on Haar invariant integrals. It takes into account the movements the robot can do in a room and the omni image transformations thus produced. A comparison with existing methods is presented using the Fisher criterion. Our method appears to get significantly better results for place recognition and robot localization, reducing in a positive way the perceptual aliasing. [ABSTRACT FROM AUTHOR]

In the recent times, many computer vision supported e-learning applications have been constructed, to provide the participants with the automated and real-time camera control capabilities. In this paper, we describe a portable and single-PC based instructional video capture system, which incorporates a variety of computer vision techniques for its video directing and close-up region specification. We describe the technologies used, including the laser-pointer detections, instructor's lip tracking and individual teaching object recognition. As the same time, we also explain how we have achieved both low-cost and portability property in our design. [ABSTRACT FROM AUTHOR]

Most existing cognitive architectures integrate computer vision and symbolic reasoning. However, there is still a gap between low-level scene representations (signals) and abstract symbols. Manually attaching, i.e. grounding, the symbols on the physical context makes it impossible to expand system capabilities by learning new concepts. This paper presents a visual bootstrapping approach for the unsupervised symbol grounding. The method is based on a recursive clustering of a perceptual category domain controlled by goal acquisition from the visual environment. The novelty of the method consists in division of goals into the classes of parameter goal, invariant goal and context goal. The proposed system exhibits incremental learning in such a manner as to allow effective transferable representation of high-level concepts. [ABSTRACT FROM AUTHOR]

Scene understanding is an important problem in intelligent robotics. Since visual information is uncertain due to several reasons, we need a novel method that has robustness to the uncertainty. Bayesian probabilistic approach is robust to manage the uncertainty, and powerful to model high-level contexts like the relationship between places and objects. In this paper, we propose a context-based Bayesian method with SIFT for scene understanding. At first, image pre-processing extracts features from vision information and objects-existence information is extracted by SIFT that is rotation and scale invariant. This information is provided to Bayesian networks for robust inference in scene understanding. Experiments in complex real environments show that the proposed method is useful. [ABSTRACT FROM AUTHOR]

This paper presents a practical combination of image processing and pattern recognition techniques in order to identify pathological and atypical cells in phase contrast cytological images. The algorithms involved in the processing cover: oriented edge detection, ridge following, contour grouping and ellipse fitting. The Hough Transform and other techniques are discussed for comparison. Various pattern recognition techniques are tested and compared. All the exploited algorithms were customized to reflect specificity of phase contrast images and apriori-knowledge of cytological smear. Possible applications of this algorithm for automated screening systems are enumerated. [ABSTRACT FROM AUTHOR]

The analysis of the events taking place in a crossroads offers the opportunity to avoid harmful situations and the potential to increase traffic efficiency in modern urban areas. This paper presents an automatic visual system that reasons about the moving vehicles being observed and extracts high-level information, useful for traffic monitoring and detection of unusual activity. Initially, moving objects are detected using an adaptive background image model. Then, the vehicles are tracked down by an iterative method where the features being tracked are updated frame by frame. Next, paths are packed into routes using a similarity measure and a sequential clustering algorithm. Finally, the crossroads activity is organized into states representing the underlying mechanism that causes the type of motion being detected. We present the experimental evidence that suggests that the framework may prove to be useful as a tool to monitor traffic-light-controlled crossroads. [ABSTRACT FROM AUTHOR]

In this paper, we propose a novel stochastic attributed relational graph (SARG) matching algorithm in order to cope with possible distortions due to noise and occlusion. The support flow and the correspondence measure between nodes are defined and estimated by analyzing the distribution of the attribute vectors in the relation vector space. And then the candidate subgraphs are extracted and ordered according to the correspondence measure. Missing nodes for each candidates are identified by the iterative voting scheme through an error analysis, and then the final subgraph matching is carried out effectively by excluding them. Experimental results on the synthetic ARGs demonstrate that the proposed SARG matching algorithm is quite robust and efficient even in the noisy environment. Comparative evaluation results also show that it gives superior performance compared to other conventional graph matching approaches. [ABSTRACT FROM AUTHOR]

Understanding behavior of migrating cells is becoming an emerging research area with many important applications. Segmentation and tracking constitute vital steps of this research. In this paper, we present an automated cell segmentation and tracking system designed to study migration of cells imaged with a phase contrast microscope. For segmentation the system uses active contour level set methods with a novel extension that efficiently prevents false-merge problem. Tracking is done by resolving frame to frame correspondences between multiple cells using a multi-distance, multi-hypothesis algorithm. Cells that move into the field-of-view, arise from cell division or disappear due to apoptosis are reliably segmented and tracked by the system. Robust tracking of cells, imaged with a phase contrast microscope is a challenging problem due to difficulties in segmenting dense clusters of cells. As cells being imaged have vague borders, close neighboring cells may appear to merge. These false-merges lead to incorrect trajectories being generated during the tracking process. Current level-set based approaches to solve the false-merge problem require a unique level set per object (the N-level set paradigm). The proposed approach uses evidence from previous frames and graph coloring principles and solves the same problem with only four level sets for any arbitrary number of similar objects, like cells. [ABSTRACT FROM AUTHOR]

A novel multiresolution approach is presented to segment Brain MRI images using fuzzy clustering. This approach is based on the fact that the image segmentation results should be optimized simultaneously in different scales. A new fuzzy inter-scale constraint based on antistrophic diffusion linkage model is introduced, which builds an efficient linkage relationship between the high resolution images and low resolution ones. Meanwhile, this paper develops two new fuzzy distances and then embeds them into the fuzzy clustering algorithm. The distances describe the fuzzy similarity in adjacent scales effectively. Moreover, a new multiresolution framework combining the inter- and intra-scale constraints is presented. The proposed framework is robust to noise images and low contrast ones, such as medical images. Segmentation of a number of images is illustrated. The experiments show that the proposed approach can extract the objects accurately. [ABSTRACT FROM AUTHOR]

In this paper, we propose a segmentation method for an automated differential counter using image analysis. The segmentation here is to extract leukocytes (white blood cells) and separate its constituents, nucleus and cytoplasm, in blood smear images. For this purpose, a region-based active contour model is used where region information is estimated using a statistical analysis. The role of the regional statistics is mainly to attract evolving contours toward the boundaries of leukocytes, avoiding problems with initialization. And contour deformation near to the boundaries is constrained by an additional regularizer. The active contour model is implemented using a level set method and validated with a leukocyte image database. [ABSTRACT FROM AUTHOR]

The paper describes a system for reading embossed Braille patterns on used aluminum beverage container lids. The intent of the system is to check whether the used containers are entitled to a refund. The lids have strong specular reflections. The reflections are avoided by a novel method that illuminates the lid alternating from two angles, and acquires two separate images. This illumination method is more compact than existing methods. We use the extended maxima algorithm to detect the Braille dots, and a cluster-based pattern point matching algorithm to recognize a pre-defined Braille pattern. The algorithms are customized to increase speed using a priori information. The system was evaluated on a test set containing 225 images. The median time used for analyzing one beverage can was 1 second, and the recognition rate was 94 percent. [ABSTRACT FROM AUTHOR]

Human interaction is a crucial restriction of active contour model, or snakes. In this paper we propose a fully automatic algorithm based on gradient vector flow (GVF) field and watershed-based region merging. Firstly a scalar force field is constructed by minimizing an energy function from the GVF force field. From the scalar field we extract a set of seed points facilely, and get an initial segmentation without doing curve evolution. Then a Region Adjacency Graph (RAG) based region merging algorithm is applied to get the final result. Several experimental results demonstrate that this method is efficient to multiple objects segmentation, and insensitive to noises. [ABSTRACT FROM AUTHOR]

In this paper, we propose a denoising method based on PCA reconstruction for removing complex color noise components on human faces, which is not easy to remove by using vectorial color filters. The proposed method is composed of the following six steps: training of canonical eigenface space using PCA, automatic extraction of facial features using active appearance model and alignment of the input face to mean shape, reconstruction of an initial noise free face, relighting of reconstructed face using a bilateral filter, extraction of noise regions using the variances of skin color of training data, and reconstruction using partial information of input images (except the noise regions) and blending of the reconstructed image with the original image. Experimental results show that the proposed denoising method maintains the structural characteristics of input faces, while efficiently removing noise components with complex colors. [ABSTRACT FROM AUTHOR]