Line segment detection is a fundamental procedure in computer vision, pattern recognition, and image analysis applications. The paper proposes a novel method for wide line segment detection especially endpoints determination based on the Guided Scale Space Radon Transform and Hessian orientations. The method begins by determining the centerlines of wide lines and then exploit the image Hessian orientations around these lines to define binary region support of the line segments and then detect endpoints. The method shows to be robust against blur and noise on synthetic images where, the evaluation of the outcomes reveals the correctness of the detection by achieving low errors. In addition, results on real images are very promising.

This paper presents a discriminative classifier for compositional data. This classifier is based on the posterior distribution of the Generalized Dirichlet which is the discriminative counterpart of Generalized Dirichlet mixture model. Moreover, following the mixture of experts paradigm, we proposed a hierarchical mixture of this classifier. In order to learn the models parameters, we use a variational approximation by deriving an upper-bound for the Generalized Dirichlet mixture. To the best of our knownledge, this is the first time this bound is proposed in the literature. Experimental results are presented for spam detection and color space identification.

In this paper, we establish the links between the Lehmer and Holder families of means and maximum weighted ¨ likelihood estimator. Considering the regular one-parameter exponential family of probability density functions, we show that the maximum weighted likelihood of the parameter is a generalized weighted mean from which Lehmer and Holder ¨ families of means are derived. Some of the outcomes obtained provide a probabilistic interpretation of these two families and could therefore broaden their uses in various applications.

Abstract An extension of Radon transform by using a measure function capturing the user need is proposed. The new transform, called scale space Radon transform, is devoted to the case where the embedded shape in the image is not filiform. A case study is brought on a straight line and an ellipse where the SSRT behaviour in the scale space and in the presence of noise is deeply analyzed. In order to show the effectiveness of the proposed transform, the experiments have been carried out, first, on linear and elliptical structures generated synthetically subjected to strong altering conditions such blur and noise and then on structures images issued from real‐world applications such as road traffic, satellite imagery and weld X‐ray imaging. Comparisons in terms of detection accuracy and computational time with well‐known transforms and recent work dedicated to this purpose are conducted, where the proposed transform shows an outstanding performance in detecting the above‐mentioned structures and targeting accurately their spatial locations even in low‐quality images.

Holes, tunnels and cavities of two‐dimensional (2D) and 3D objects are concise topological features used for object representation and recognition. In this study, the authors are representing any cubical tessellation (regular or not) of 2D and 3D objects and dealing with the extraction and the localisation of these features by using homology‐based approach. The cubical tessellation (regular or not) of objects is translated into algebraic language suitable for building a reduced cell complex structure. The extraction of the homology information is equivalent to the estimation of the rank of the homology groups of the reduced complex. The localisation means the reconstruction of the object cycles from the generators of the homology groups. The reduction operation of the cell complex leads to an efficient algorithm. Note that, several objects can be analysed simultaneously by the algorithm conceived in our approach. This algorithm is validated by using 2D and 3D binary images.

It is known that k-means clustering is especially sensitive to initial starting centers. In this paper, we propose an original version of k-means for the segmentation of Terahertz images, called ranked-k-means, which is essentially less sensitive to the initialization of the centers. We present the ranked set sampling design and explain how to reformulate the k-means technique under the ranked sample to estimate the expected centers as well as the clustering of the observed data. Our clustering approach is tested on various Terahertz images. Experimental results show that k-means based on the ranked sample is more efficient than other clustering techniques.

Even with the advancing laser scanner technology, the digital objects are inevitably corrupted with random noise during the acquisition process. Many algorithms, regardless of principle, share the same basic idea of noise reduction through mesh smoothing. Smoothing can be performed locally, as in the anisotropic filtering by calculus of variations; or by smoothing mesh attributes such as normal vector and then adjusting vertex positions. In this study, the authors propose a novel algorithm based on the statistical distribution noise for anisotropic mesh smoothing. In this approach, the probability density function of the noise is first estimated, and then the anisotropic smoothing operation is performed through the diffusion tensor that allows to preserve mesh characteristics like edges, corner and ridges. The proposed algorithm is easy to implement and is thus suitable for real‐time noise reduction application.

In this paper, a diffusion-based generic demosaicing algorithm is proposed which can be used for various sensor images captured by digital cameras equipped with various RGB color filter arrays. This algorithm improves our previous edge-sensing generic demosaicing algorithm by enhancing the computation of the green band. In fact, since the green band plays a major and crucial role in the performance of the edge-sensing generic demosaicing algorithm, a diffusion-based model is used for reducing the errors generated when computing the green band. A series of tests has been made on images of the Kodak database, and our diffusion-based demosaicing algorithm performs better than the edge-sensing generic demosaicing algorithm in regard to both subjective and objective evaluation.

We introduce a novel hierarchical mixture model where each component is composed of a set of finite probability densities forming a super class mixture. Our proposed model can be viewed as a mixture of mixtures to support multi-level hierarchies where the structure of the hierarchy can be altered according to users' ontological models within costless computational time. The proposed approach is generalized to adopt any probability density function and an algorithm to learn the model is proposed. In this paper, we adopt the inverted Dirichlet distribution to build the model, and a simulation study is performed to validate the proposed approach using synthetic and a real world challenging application concerning visual scenes categorization.

In this study, the authors analyse two well‐known image quality metrics, peak‐signal‐to‐noise ratio (PSNR) as well as structural similarity index measure (SSIM), and the authors derive an analytical relationship between them which works for some kinds of common image degradations such as Gaussian blur, additive Gaussian noise, Jpeg and Jpeg2000 compressions. The analytical relationship brings more clarity on the interpretation of PSNR and SSIM values, explains some differences found between these quality measures in the literature and confirms some experimental observations regarding these measures. A series of tests realised on images from the Kodak database give a better understanding of the performance of SSIM and PSNR in assessing image quality.

Clustering has been a subject of extensive research in data mining, pattern recognition, and other areas for several decades. The main goal is to assign samples, which are typically non-Gaussian and expressed as points in high-dimensional feature spaces, to one of a number of clusters. It is well known that in such high-dimensional settings, the existence of irrelevant features generally compromises modeling capabilities. In this paper, we propose a variational inference framework for unsupervised non-Gaussian feature selection, in the context of finite generalized Dirichlet (GD) mixture-based clustering. Under the proposed principled variational framework, we simultaneously estimate, in a closed form, all the involved parameters and determine the complexity (i.e., both model an feature selection) of the GD mixture. Extensive simulations using synthetic data along with an analysis of real-world data and human action videos demonstrate that our variational approach achieves better results than comparable techniques.

In this paper, we formalize content-based image suggestion (CBIS) as a Bayesian prediction problem. In CBIS, users provide the rating of images according to both their long-term needs and the contextual situation, such as time and place, to which they belong. Therefore, a CBIS model is defined to fit the distribution of the data in order to predict relevant images for a given user. Generally, CBIS becomes challenging when only a small amount of data is available such as in the case of "new users" and "new images." The Bayesian predictive approach is an effective solution to such a problem. In addition, this approach offers efficient means to select highly rated and diversified suggestions in conformance with theories in consumer psychology. Experiments on a real data set show the merits of our approach in terms of image suggestion accuracy and efficiency.

In this paper, we focus on the variational learning of finite Dirichlet mixture models. Compared to other algorithms that are commonly used for mixture models (such as expectation-maximization), our approach has several advantages: first, the problem of over-fitting is prevented; furthermore, the complexity of the mixture model (i.e., the number of components) can be determined automatically and simultaneously with the parameters estimation as part of the Bayesian inference procedure; finally, since the whole inference process is analytically tractable with closed-form solutions, it may scale well to large applications. Both synthetic and real data, generated from real-life challenging applications namely image databases categorization and anomaly intrusion detection, are experimented to verify the effectiveness of the proposed approach.

In recent years, an increasing number of security threats have brought a serious risk to the internet and computer networks. Intrusion Detection System (IDS) plays a vital role in detecting various kinds of attacks. Developing adaptive and flexible oriented IDSs remains a challenging and demanding task due to the incessantly appearance of new types of attacks and sabotaging approaches. In this paper, we propose a novel unsupervised statistical approach for detecting network based attacks. In our approach, patterns of normal and intrusive activities are learned through finite generalized Dirichlet mixture models, in the context of Bayesian variational inference. Under the proposed variational framework, the parameters, the complexity of the mixture model, and the features saliency can be estimated simultaneously, in a closed-form. We evaluate the proposed approach using the popular KDD CUP 1999 data set. Experimental results show that this approach is able to detect many different types of intrusions accurately with a low false positive rate.

In this paper, we introduce a new algorithm that can be used for reducing aliasing in images. Our algorithm, which is derived from the standard diffusion equation, cancels the aliasing of step edges found in images by reducing their curvature while preserving their contrast through a high-pass filter. Our algorithm can be seen as an adaptive level-curve method in which diffusion is carried out in the normal direction of the gradient. Experimental tests based on different grey-level images show that our algorithm efficiently reduces aliasing, which is confirmed through objective and subjective quality evaluations.

In this paper, we introduce a new demosaicing algorithm that can be used for various sensor images captured by digital cameras equipped with various red-green-blue color filter arrays. Our algorithm enhances the universal demosaicing algorithm of Lukac et al by defining a new spectral interpolation model that exploits not only the information on the color of pixels but also the relative distance between neighboring pixels within an image. Moreover, we include an edge-detection model that makes our algorithm adaptive and reduces the presence of color shifts and artifacts. A series of tests has been made on images of the Kodak database, and our algorithm performs better than the universal demosaicing algorithm with regard to both subjective and objective evaluation. The versatility of our demosaicing algorithm is also highlighted through an application to the issue of color image resampling, and we obtain conclusive experimental results.

In this paper, we analyse two well-known objective image quality metrics, the peak-signal-to-noise ratio (PSNR) as well as the structural similarity index measure (SSIM), and we derive a simple mathematical relationship between them which works for various kinds of image degradations such as Gaussian blur, additive Gaussian white noise, jpeg and jpeg2000 compression. A series of tests realized on images extracted from the Kodak database gives a better understanding of the similarity and difference between the SSIM and the PSNR.

In this paper, the Radon transform is used to design a new descriptor called Phi-signature invariant to usual geometric transformations. Experiments show the effectiveness of the multilevel representation of the descriptor built from Phi-signature and R-signature, compared to the powerful generic Fourier descriptor.

In this paper, a parametric and unsupervised histogram-based image segmentation method is presented. The histogram is assumed to be a mixture of asymmetric generalized Gaussian distributions. The mixture parameters are estimated by using the Expectation Maximization algorithm. Histogram fitting and region uniformity measures on synthetic and real images reveal the effectiveness of the proposed model compared to the generalized Gaussian mixture model.

In this letter, we propose a clustering model that efficiently mitigates image and video under/over-segmentation by combining generalized Gaussian mixture modeling and feature selection. The model has flexibility to accurately represent heavy-tailed image/video histograms, while automatically discarding uninformative features, leading to better discrimination and localization of regions in high-dimensional spaces. Experimental results on a database of real-world images and videos showed us the effectiveness of the proposed approach.

In this paper, we propose a clustering algorithm based on both Dirichlet processes and generalized Dirichlet distribution which has been shown to be very flexible for proportional data modeling. Our approach can be viewed as an extension of the finite generalized Dirichlet mixture model to the infinite case. The extension is based on nonparametric Bayesian analysis. This clustering algorithm does not require the specification of the number of mixture components to be given in advance and estimates it in a principled manner. Our approach is Bayesian and relies on the estimation of the posterior distribution of clusterings using Gibbs sampler. Through some applications involving real-data classification and image databases categorization using visual words, we show that clustering via infinite mixture models offers a more powerful and robust performance than classic finite mixtures.

This paper presents an unsupervised approach for feature selection and extraction in mixtures of generalized Dirichlet (GD) distributions. Our method defines a new mixture model that is able to extract independent and non-Gaussian features without loss of accuracy. The proposed model is learned using the Expectation-Maximization algorithm by minimizing the message length of the data set. Experimental results show the merits of the proposed methodology in the categorization of object images.

We address the problem of modeling and adapting JPEG to the energy requirements of visual sensor networks (VSN). For JPEG modeling purposes, we develop a simplified high-level energy consumption model for each stage of JPEG-like scheme, which can be used to roughly evaluate the energy dissipated by a given visual sensor. This model is based on the basic operations needed at each stage of JPEG, and it does not take into account the complexity of implementation. For JPEG adaptation, we propose to process only a reduced part of each block of 8times8 DCT coefficients of the target image, which minimizes the dissipated energy and maximizes the system lifetime, while preserving an adequate image quality at the sink.

Image segmentation combining boundary and region information has been the subject of numerous research works in the past. This combination is usually subject to arbitrary weighting parameters (hyper-parameters) that control the contribution of boundary and region features during segmentation. In this work, we investigate a new approach for estimating the hyper-parameters adaptively to segmentation. The approach takes its roots from the physical properties of the energy functional controlling segmentation and a Bayesian formulation of segmentation and hyper-parameters estimation.

With Visual Sensor Networks (VSN), designers must respect strict constraints on energy consumption, which make compression standards, such as JPEG, not energy-beneficial to VSN. Our approach for tackling this constraint problem consists in adapting JPEG by exploiting the DCT energy compaction property. This exploitation is performed by processing only a portion of each block of 8 times 8 DCT coefficients of the captured image. This approach induces two conflicting effects. Indeed, reducing the size of the portion of DCT block presents the advantage of reducing the energy consumed for processing and transmitting an image, but it also presents the drawback of reducing the quality of the image received at the sink.We propose two methods to solve this conflict: a global method and a local method. In the global method, an optimal size is computed for all portions of DCT blocks of a whole image.

In this paper, we propose a manifold smoothing method based on the heat diffusion process. We start from the global equation of heat conservation and we decompose it into basic laws. The numerical scheme is derived in a straightforward way from the discretization of the basic heat transfer laws using computation algebraic topological tools CAT, thus providing a physical and topological explanation for each step of the discretization process.

Abstract In this paper, we present a probabilistic approach for 3D object's smoothing. The core idea behind the proposed method is to relate the problem of smoothing objects to that of tracking the transition probability density functions of an underlying random process. We show that such an approach allows for additional insight and sufficient flexibility compared with existing standard smoothing techniques. In particular, we are able to propose a newer, faster, and simpler smoothing approach that retains and enhances important manifold features. Furthermore, it is demonstrated to improve performance over existing smoothing techniques.

In this paper, we investigate the effectiveness of a Bayesian logistic regression model to compute the weights of a pseudo-metric, in order to improve its discriminatory capacity and thereby increase image retrieval accuracy. In the proposed Bayesian model, the prior knowledge of the observations is incorporated and the posterior distribution is approximated by a tractable Gaussian form using variational transformation and Jensen's inequality, which allow a fast and straightforward computation of the weights. The pseudo-metric makes use of the compressed and quantized versions of wavelet decomposed feature vectors, and in our previous work, the weights were adjusted by classical logistic regression model. A comparative evaluation of the Bayesian and classical logistic regression models is performed for content-based image retrieval as well as for other classification tasks, in a decontextualized evaluation framework. In this same framework, we compare the Bayesian logistic regression model to some relevant state-of-the-art classification algorithms. Experimental results show that the Bayesian logistic regression model outperforms these linear classification algorithms, and is a significantly better tool than the classical logistic regression model to compute the pseudo-metric weights and improve retrieval and classification performance. Finally, we perform a comparison with results obtained by other retrieval methods.

Existing recommender systems provide an elegant solution to the information overload in current digital libraries such as the Internet archive. Nowadays, the sensors that capture the user's contextual information such as the location and time are become available and have raised a need to personalize recommendations for each user according to his/her changing needs in different contexts. In addition, visual documents have richer textual and visual information that was not exploited by existing recommender systems. In this paper, we propose a new framework for context-aware recommendation of visual documents by modeling the user needs, the context and also the visual document collection together in a unified model. We address also the user's need for diversified recommendations. Our pilot study showed the merits of our approach in content based image retrieval.

Most image segmentation algorithms in the past are based on optimizing an objective function that aims to achieve the similarity between several low-level features to build a partition of the image into homogeneous regions. In the present paper, we propose to incorporate the relevance (selection) of the grouping features to enforce the segmentation toward the capturing of objects of interest. The relevance of the features is determined through a set of positive and negative examples of a specific object defined a priori by the user. The calculation of the relevance of the features is performed by maximizing an objective function defined on the mixture likelihoods of the positive and negative object examples sets. The incorporation of the features relevance in the object segmentation is formulated through an energy functional which is minimized by using level set active contours. We show the efficiency of the approach on several examples of object of interest segmentation and tracking where the features relevance is used.

Many computer vision and pattern recognition problems involve the use of finite Gaussian mixture models. Finite mixture model using generalized Dirichlet distribution has been shown as a robust alternative of normal mixtures. In this paper, we adopt a Bayesian approach for generalized Dirichlet mixture estimation and selection. This approach, offers a solid theoretical framework for combining both the statistical model learning and the knowledge acquisition. The estimation of the parameters is based on the Monte Carlo simulation technique of Gibbs sampling mixed with a Metropolis-Hastings step. For the selection of the number of clusters, we used Bayes factors. We have successfully applied the proposed Bayesian framework to model IR eyes. Experimental results are shown to demonstrate the robustness, efficiency, and accuracy of the algorithm.

Content-based image suggestion (CBIS) targets the recommendation of products based on user preferences on the visual content of images. In this paper, we mo- tivate both feature selection and model order identification as two key issues for a successful CBIS. We propose a generative model in which the visual features and users are clustered into separate classes. We identify the number of both user and image classes with the simultaneous selection of relevant visual features us- ing the message length approach. The goal is to ensure an accurate prediction of ratings for multidimensional non-Gaussian and continuous image descriptors. Experiments on a collected data have demonstrated the merits of our approach.

We consider the problem of determining the structure of high-dimensional data, without prior knowledge of the number of clusters. Data are represented by a finite mixture model based on the generalized Dirichlet distribution. The generalized Dirichlet distribution has a more general covariance structure than the Dirichlet distribution and offers high flexibility and ease of use for the approximation of both symmetric and asymmetric distributions. This makes the generalized Dirichlet distribution more practical and useful. An important problem in mixture modeling is the determination of the number of clusters. Indeed, a mixture with too many or too few components may not be appropriate to approximate the true model. Here, we consider the application of the minimum message length (MML) principle to determine the number of clusters. The MML is derived so as to choose the number of clusters in the mixture model which best describes the data. A comparison with other selection criteria is performed. The validation involves synthetic data, real data clustering, and two interesting real applications: classification of web pages, and texture database summarization for efficient retrieval.

In this paper, we present a new framework for organizing image collections into structures that can be used for indexing, browsing, retrieval and summarization. Instead of using tree-based techniques which are not suitable for images, we develop a new solution that is specifically designed for image collections. We consider both low-level image content and high-level semantics in an attempt to alleviate the semantic gap encountered by many systems. The fact that our model is based on a probabilistic framework makes it possible to combine it in a natural way with probabilistic techniques developed recently for image retrieval. The structure our model generates is applied for four purposes. The first is to provide retrieval module with an index, which allows it to improve retrieval time and accuracy, while the second is to provide users with a hierarchical browsing catalog that allows them to navigate the image collection by subject. This represents an additional step towards facilitating human-computer interaction in the context of image retrieval and navigation. The third aim is to provide users with a summarization of the general content of each class in the collection, and the fourth is a retrieval mechanism. Related issues such as relevance feedback and feature selection are also addressed. The experiments at the end of the paper show that the proposed framework yields some significant improvements

In this paper, we propose a novel object tracking algorithm in video sequences. The method is based on object mixture matching between successive frames of the sequence by using active contours. Only the segmentation of the objects in the first frame is required for initialization. The evolution of the object contour on a current frame aims to find the maximum fidelity of the mixture likelihood for the same object between successive frames while having the best fit of the mixture parameters to the homogenous parts of the objects. To permit for a precise and robust tracking, region, boundary and shape information are coupled in the model. The method permits for tracking multi-class objects on cluttered and non-static backgrounds. We validate our approach on examples of tracking performed on real video sequences.

This paper proposes an alternative to partial differential equations (PDEs) for solving problems in computer vision based on evolutive heat transfer. Traditionally, the method for solving such physics-based problems is to discretize and solve a PDE by a purely mathematical process. Instead of using the PDE, we propose to use the global heat principle and to decompose it into basic laws. We show that some of these laws admit an exact global version since they arise from conservative principles. We also show that the assumptions made about the other basic Iaws can be made wisely, taking into account knowledge about the problem and the domain. The numerical scheme is derived in a straightforward way from the modeled problem, thus providing a physical explanation for each step in the solution. The advantage of such an approach is that it minimizes the approximations made during the whole process and it modularizes it, allowing changing the application to a great number of problems. We apply the scheme to two applications: image denoising and inpainting which are modeled with heat transfer. For denoising, we propose a new approximation for the conductivity coefficient and we add thin lines to the features in order to block diffusion.

This paper applies a robust statistical scheme to the problem of unsupervised learning of high-dimensional data. We develop, analyze, and apply a new finite mixture model based on a generalization of the Dirichlet distribution. The generalized Dirichlet distribution has a more general covariance structure than the Dirichlet distribution and offers high flexibility and ease of use for the approximation of both symmetric and asymmetric distributions. We show that the mathematical properties of this distribution allow high-dimensional modeling without requiring dimensionality reduction and, thus, without a loss of information. This makes the generalized Dirichlet distribution more practical and useful. We propose a hybrid stochastic expectation maximization algorithm (HSEM) to estimate the parameters of the generalized Dirichlet mixture. The algorithm is called stochastic because it contains a step in which the data elements are assigned randomly to components in order to avoid convergence to a saddle point. The adjective "hybrid" is justified by the introduction of a Newton-Raphson step. Moreover, the HSEM algorithm autonomously selects the number of components by the introduction of an agglomerative term. The performance of our method is tested by the classification of several pattern-recognition data sets. The generalized Dirichlet mixture is also applied to the problems of image restoration, image object recognition and texture image database summarization for efficient retrieval. For the texture image summarization problem, results are reported for the Vistex texture image database from the MIT Media Lab.

In content-based image retrieval, understanding the user's needs is a challenging task that requires integrating him in the process of retrieval. Relevance feedback (RF) has proven to be an effective tool for taking the user's judgement into account. In this paper, we present a new RF framework based on a feature selection algorithm that nicely combines the advantages of a probabilistic formulation with those of using both the positive example (PE) and the negative example (NE). Through interaction with the user, our algorithm learns the importance he assigns to image features, and then applies the results obtained to define similarity measures that correspond better to his judgement. The use of the NE allows images undesired by the user to be discarded, thereby improving retrieval accuracy. As for the probabilistic formulation of the problem, it presents a multitude of advantages and opens the door to more modeling possibilities that achieve a good feature selection. It makes it possible to cluster the query data into classes, choose the probability law that best models each class, model missing data, and support queries with multiple PE and/or NE classes. The basic principle of our algorithm is to assign more importance to features with a high likelihood and those which distinguish well between PE classes and NE classes. The proposed algorithm was validated separately and in image retrieval context, and the experiments show that it performs a good feature selection and contributes to improving retrieval effectiveness.

This paper proposes an unsupervised algorithm for learning a finite Dirichlet mixture model. An important part of the unsupervised learning problem is determining the number of clusters which best describe the data. We extend the minimum message length (MML) principle to determine the number of clusters in the case of Dirichlet mixtures. Parameter estimation is done by the expectation-maximization algorithm. The resulting method is validated for one-dimensional and multidimensional data. For the one-dimensional data, the experiments concern artificial and real SAP image histograms. The validation for multidimensional data involves synthetic data and two real applications: shadow detection in images and summarization of texture image databases for efficient retrieval. A comparison with results obtained for other selection criteria is provided

In this paper, we propose a novel automatic framework for variational color image segmentation based on unifying adaptive region information and mixture modelling. We consider a formulation of the region information based on the posterior probability of a mixture of general Gaussian (GG) pdfs where each region is represented by a pdf. The segmentation is formulated by the minimization of an energy functional according to the region contours and all the mixture parameters respectively. Two main objectives are achieved by the approach. A scheme is provided to extend easily the adaptive segmentation to an arbitrary number of regions and to perform it in a fully automatic fashion. Moreover, the segmentation recovers an accurate and representative mixture of pdfs. In the approach, we couple the boundary and region information of the image to steer the segmentation. We validate the method on the segmentation of real world color images.

The performance of a statistical image processing system depends in large part on the accuracy of the probabilistic model used. This paper presents a robust probabilistic mixture model based on the Dirichlet distribution. An unsupervised algorithm based on MML for learning this mixture is given, too. Experimental results involve shadows modeling and its application to shadows detection in images.

There are many image fusion methods that can be used to produce high-resolution multispectral images from a high-resolution panchromatic image and low-resolution multispectral images. Starting from the physical principle of image formation, this paper presents a comprehensive framework, the general image fusion (GIF) method, which makes it possible to categorize, compare, and evaluate the existing image fusion methods. Using the GIF method, it is shown that the pixel values of the high-resolution multispectral images are determined by the corresponding pixel values of the low-resolution panchromatic image, the approximation of the high-resolution panchromatic image at the low-resolution level. Many of the existing image fusion methods, including, but not limited to, intensity-hue-saturation, Brovey transform, principal component analysis, high-pass filtering, high-pass modulation, the a/spl grave/ trous algorithm-based wavelet transform, and multiresolution analysis-based intensity modulation (MRAIM), are evaluated and found to be particular cases of the GIF method. The performance of each image fusion method is theoretically analyzed based on how the corresponding low-resolution panchromatic image is computed and how the modulation coefficients are set. An experiment based on IKONOS images shows that there is consistency between the theoretical analysis and the experimental results and that the MRAIM method synthesizes the images closest to those the corresponding multisensors would observe at the high-resolution level.

The performance of a statistical signal processing system depends in large part on the accuracy of the probabilistic model used. This paper presents a robust probabilistic mixture model based on a generalization of the Dirichlet distribution. An unsupervised algorithm for learning this mixture is given, too. The proposed approach for estimating the parameters of a Dirichlet mixture is based on the maximum likelihood (ML) and Fisher scoring methods. Experimental results involve human skin color modeling and its application to skin detection in images.

The problem of recovering two primitive images from their transparent combination is explored. Given an image sequence, the case where each primitive image undergoes a dissimilar and invertible motion over time is considered. The authors show that discrete samples of the primitive images can be recovered from the observation of two consecutive image frames. In this context, consideration is given to the density of the recovered sample set and the requirements for proper reconstruction. On the whole, transparency separation and recovery is an ill-posed inverse problem in the sense that its solution is not unique. Recovery, however, can be successfully performed by finding the smoothest solution consistent with the data. Illustrative examples show that the proposed method provides good estimates of the primitive images.

We present a relevance feedback model for CBIR, based on a feature weighting algorithm. The proposed model uses positive and negative items selected by the user to learn the importance of image features, then applies the obtained weights to define similarity measures corresponding to the user's perception. The basic principle of this work is to give more importance to features with a high likelihood and those which separate well between positive example (PE) classes and negative example (NE) classes. The proposed algorithm was validated separately and in the image retrieval context, and the experiments show that it contributes in improving retrieval effectiveness.

This paper presents an unsupervised algorithm for learning a finite mixture model from multivariate data. This mixture model is based on the Dirichlet distribution, which offers high flexibility for modeling data. The proposed approach for estimating the parameters of a Dirichlet mixture is based on the maximum likelihood (ML) and Fisher scoring methods. Experimental results are presented for the following applications: estimation of artificial histograms, summarization of image databases for efficient retrieval, and human skin color modeling and its application to skin detection in multimedia databases.

This paper presents a homotopy-based algorithm for simultaneous recovery of defocus blur and the affine transformation between two images of the same scene. One of the images (and its partial derivatives) is expressed as a function of the second image, partial derivatives of the two images, blur difference, affine parameters and a continuous parameter derived from homotopy methods. All of these unknowns can thus be directly computed by resolving a system of equations. The proposed algorithm is tested using synthetic and real images. The results confirm that dense and accurate estimation can be obtained.

This paper addresses the problem of image combination. The mathematical model of a general image combination (GIC) method is driven from the physical principle of image formation. It is shown that many existing image combination methods are the particular cases of the GIC method. The performance of various image combination methods is then analyzed based on the advantages and disadvantages of different assumptions.

The paper presents an approach for roads detection based on synthetic aperture radar (SAR) images and road databases. The vectors provided by the database are refined using active contours (snakes). In this framework, we firstly develop a restoration filter based on the frost filter achieving an acceptable compromise between speckle elimination and lines preserving. This is followed by a line plausibility calculation step which is used to deform the snake from its initial location toward the final solution. The snake is reformulated using finite elements method. The setting of the snake parameters is not an obvious problem especially when they are tuned by trial-and-error process. We propose a new automatic computational rule for the snake parameters. Our approach is validated by a series of tests on synthetic and SAR images.

Texture is a very important image feature extremely used in various image processing problems. It has been shown that humans use some perceptual textural features to distinguish between textured images or regions. Some of the most important features are coarseness, contrast, direction and busyness. In this paper a new method based on the autocovariance function to estimate quantitatively these features is shown and the correspondence between these computational measures and the psychological ones made by human subjects is shown using some psychometric method.

It has been shown that humans use some perceptual textural features such as coarseness, contrast and direction to distinguish between textured images or regions. The aim of this paper is to present a new method to estimate these perceptual textural features using the autocovariance function. Computational measures derived from the autocovariance function to estimate these perceptual textural features are presented. Experimental results are then given and the correspondence between the computational measures proposed and the psychological measures is shown using some psychometric method.

This paper describes an efficient approach for the detection of line junctions in gray-level images. The algorithm is divided into two steps. First, given the lines extracted from the original image, local line curvature is estimated. For this purpose, two different measures of curvature are proposed: the rate of change of direction of the orientation vector along the line, and the mean of the dot products of orientation vectors within a given neighborhood. The second step involves the localization of junctions. Examples are provided based on experiments with synthetic and real images.

An optimal line detector for the one-dimensional case is derived from Canny's criteria (1986). The detector is extended to the two-dimensional case by operating separately in the x and y directions. An efficient implementation using an infinite impulse response (IIR) filter is provided. This implementation has an additional advantage that increasing the filter scale affects neither temporal nor spatial complexity. Our detector is faster than the Gaussian used by Steger (1998); e.g., when the scale is 3 our detector is 33 times faster. Experimental results using real images demonstrate the validity of the algorithm.

In this paper, we propose a phase-based disparity estimation algorithm embedded in two different cooperative multiscale schemes. The first disparity integration scheme selects the best disparity as the one having the highest magnitude of the Gabor coefficient through the scales, while the second scheme assumes that the best disparity is the one for which the local spatial frequency is the nearest, in relative distance, from the peak tuning frequency of the filter. We show that both these schemes are efficient and roughly equivalent since they give the same results.

This paper presents an algorithm for a dense computation of the difference in blur between two images. The two images are acquired by varying the intrinsic parameters of the camera. The image formation system is assumed to be passive. Estimation of depth from the blur difference is straightforward. The algorithm is based on a local image decomposition technique using the Hermite polynomial basis. We show that any coefficient of the Hermite polynomial computed using the more blurred image is a function of the partial derivatives of the other image and the blur difference. Hence, the blur difference can be computed by resolving a system of equations. All computations required are local and carried out in the spatial domain. An algorithm is presented for estimation of the blur in 1D and 2D cases and its behavior is studied for constant images, step edges, line edges and junctions. The algorithm is tested using synthetic and real images. The results obtained are very encouraging.

This paper presents an algorithm for a dense computation of the difference in blur between two images. The two images are acquired by varying the intrinsic parameters of the camera. The image formation system is assumed to be passive. The algorithm is based on a local image decomposition technique using the Hermite polynomial basis. We show that any coefficient of the Hermite polynomial computed using the unfocused image is a function of the partial derivatives of the focused image and the blur difference. Hence, the blur difference can be computed by resolving a system of equations. An algorithm is presented for estimation of the blur in 1D and 2D images. The algorithm is tested using synthetic and real images. The results obtained are very encouraging.

Content-based image retrieval is emerging as an important research area with potential applications in many domains, notably multimedia databases and digital libraries. The goal of the present research is to create a usable system that performs automatic picture labeling. Recent studies have shown that some improvements can be achieved with a system which selects the best from amongst many models of representation. The focus of our research is thus twofold: creation of models and selection of the best ones for each label.

An automatic detection model for ship targets in RADARSAT SAR images is being developed by using statistical methods, Radon transform and other image processing techniques. This paper presents current progress made on the detection model.

This paper presents a new 2D disparity-estimation algorithm. Disparity is calculated from an estimation of the phase difference between two views of the same scene from different angles. This phase difference is obtained by convolution of the two views with the Gaussian function and its first two derivatives. Since the disparities obtained often contain errors, we use a regularization model in order to smooth the disparity by preserving the structure of objects. While the models used were simple and noise-free, the results obtained are promising.

This paper concerns the influence of edge direction on the estimation of edge contrast and orientation. We show that the gradient estimated using radial filters is not affected by edge orientation. For non-radial filters the gradient can be affected by edge orientation. For instance, we find that the estimated edge orientation using a non-radial filter may be biased, even if the signal is noise-free. However, there are non-radial filters for which gradient is unaffected by edge orientation as in the case of radial filters. The properties of these functions are given in this paper. The results are illustrated by the study of the Canny, Deriche, and Shen & Castan detectors. We take into account discretization errors. These results give a clear indication of the effect of the rotation invariance property of an edge detector on its response, thus providing a more precise meaning for this property in edge detection.

This paper summarizes a system, called SED, "Selection-of Edge Detectors", which is able to automatically select edge detectors and their scales to extract a given edge. The basic organization of the SED system is a collection of edge detectors within a knowledge structure about the characteristics of the given edge, the properties of the detectors and the mutual relation between them. The combination of this information in the selection process provides a basis for avoiding a combinatorial search for appropriate edge detectors. We illustrate our approach by an example.

We address the problems of the influence of the edge orientation on the performance of first order operators. To illustrate, we present the performance evaluation, including a subpixel error of three popular detectors: Canny, Deriche, as well as Shen and Castan.

An edge combination algorithm is described. The authors' approach is based on the study of four step edge models (ideal, blurred, pulse and staircase) in scale space. Under these conditions, it is shown that the use of two scales (high and low) is sufficient for good edge detection. A set of rules is derived to combine edge information, and an appropriate algorithm is given, taking into account the origin of false edges and their behavior in scale space.>

Describes a new approach for positioning boundaries in a discrete image to subpixel values. To correct the position of step edges, the authors combine the operator output and the properties of both operator and edge. This method is simpler to use than existing ones and requires a relatively small amount of computer power. Both availability and reliability are discussed for first and second order operators.>

Summarizes a system, called SED 'selection of edge detectors', which is able to select automatically edge detectors and their scales to extract a given edge. To avoid a combinatorial approach, the authors use several sources of information like the characteristics of the given edge, the properties of the detectors and the mutual relations between them. The authors illustrate our approach by the study of an example.>