Snakes, or active contours, are one of the major paradigms in image segmentation. With the gradient vector flow (GVF) as external force, they have a larger capture range and the ability to progress into concave boundaries. However, when we have to deal with highly non-convex shapes, the GVF field forms an area where the forces point in opposite directions and the snake stops. GVF is built as a diffusion process of the gradient vectors of an edge map derived from the image. In this paper, we will view the diffusion process as a mechanism having the gradient vectors of the edge map as the input and the resulting GVF as the output. We will show how a modified input will result in a GVF able to drive the snake in such a highly non-convex boundary. This ability is revealed as a generalized property of the GVF diffusion process.

The extension of the standard grayscale active appearance model (AAM) techniques to color images is investigated. Prior work in this field has mainly focused on RGB color models which did not demonstrate noticeable benefits over grayscale models from the point of view of convergence accuracy. We improve on previous work by normalizing the color texture vector separately for intensity and chromaticity components. Where an appropriate color space is chosen, we demonstrate improvements in convergence accuracy as well as image synthesization quality for AAMs. Optimal results are achieved when a color space in which the image channels are strongly decorrelated is chosen. Our best results are achieved using the I1I2I3 color space, originally proposed by Ohta.

Synthetic aperture radar (SAR) interferometric data offer the opportunity to measure temperate glacier surface topography and displacement. The increase of the resolution provided by the most recent SAR systems has some critical implications. For instance, a reliable estimate of the phase gradient can only be achieved by using interferogram local frequencies. In this paper, an original two-step method for estimating local frequencies is proposed. The 2-D phase signal is considered to have two deterministic components corresponding to low-resolution (LR) fringes and high-resolution (HR) patterns due to the local microrelief, respectively. The first step of the proposed algorithm consists in the LR phase flattening. In the second step, the local HR frequencies are estimated from the phase 2-D autocorrelation function computed on adaptive neighborhoods. This neighborhood is the set of connected pixels belonging to the same HR spatial feature and respecting the ldquolocal stationarityrdquo hypothesis. Results with both simulated TerraSAR-X interferograms and real airborne E-SAR images are presented to illustrate the potential of the proposed method.

SAR interferometry (InSAR) performs two acquisitions (spatially separated by the baseline) of the signal back-scattered by the resolution cell which contains height and/or displacement information. Repeat pass spaceborne interferometry provides multi-interferograms which can be used to extract such information either by combining the multi-temporal results of conventional interferometry or by a different approach based on specific targets: the coherent stable scatterers (CSS). In this paper a two-step approach is proposed to obtain specific features from multi-temporal InSAR data sets. The first step consists in extracting image attributes related to the useful information. The second step consists in merging the attributes using an interactive fuzzy fusion technique. The interactive fuzzy fusion is proposed to provide end-users with a simple and easily understandable tool for tuning the detection results. The method is applied on a data set of five co-registered ERS 1/2 tandems from the French Alps (the Mont-Blanc region), including two temperate glaciers: the Argentiere and the Mer-de-glace. The results illustrate how the end-user can combine the proposed attributes to detect the presence of CSS or distributed stable scatterers usefull for multi-temporal analysis.

In this paper an improved cut detection algorithm, adapted to the segmentation of animation movies, is proposed. As color is a major feature of animation movies (each movie has its own particular color distribution) the proposed algorithm applies second order derivatives on Euclidean distances between color histograms of frames quadrants in order to improve the cut detection. For the frame classification, an automatic threshold estimation is proposed. Also, in order to reduce false detections, we propose an algorithm to detect an effect specific to animation movies, named "short color change" (i.e. thunders, lightening). The resulting method achieved better results compared to the classical histogram-based and motion-discontinuity based approaches, as shown by tests conducted on several animation movies

In the field of medical image indexation, automatic categorization provides the means for extracting, otherwise unavailable, information from images. Our work is focused on content-based automatic medical image categorization methods, in the on-line context of the CISMeF health-catalogue. In this study we propose and evaluate a reduced symbolic image representation. The categorization of medical images according to their modality, anatomic region and view angle is based on texture and statistical features. We use a medical image dataset of 10322 images, representing 33 classes, manually annotated by an experienced radiologist. A top classification accuracy of 92.43% is obtained using k-Nearest Neighbors classifier on a 64-label symbolic representation. This shows that the compact symbolic image representation we propose conveys enough of the initial texture information to obtain high recognition rates, despite the complex context of multi-modal medical image categorization.

This paper presents a method for the automatic generation of animation movie trailers. First, the movie is divided into shots, by detecting the video transitions (cuts, fades and dissolves) and an animation movie specific color effect named short color change or SCC (i.e. explosions, thunders). The movie action content is further highlighted by analyzing the movie at two different granularity levels. First, an inter-shot analysis is performed by measuring the video transition temporal distribution. As repetitive shot changes are related to action, we define as an action shot, a movie segment containing a high shot change ratio. On the other hand, an inter-frame analysis is performed: for each shot within an action shot, an histogram of cumulative inter-frame distances is computed to serve as a measure of the frame spatial activity. Repetitive color changes are also related to action. Since movie trailers only show some of the most attractive movie scenes, the proposed trailer is a moving-image abstract computed on the retained action shots. It provides the user with a compact and efficient representation of the movie action content. The proposed approach was tested on several animation movies.

In this paper, a new method to filter coherency matrices of polarimetric or interferometric data is presented. For each pixel, an adaptive neighborhood (AN) is determined by a region growing technique driven exclusively by the intensity image information. All the available intensity images of the polarimetric and interferometric terms are fused in the region growing process to ensure the validity of the stationarity assumption. Afterward, all the pixels within the obtained AN are used to yield the filtered values of the polarimetric and interferometric coherency matrices, which can be derived either by direct complex multilooking or from the locally linear minimum mean-squared error (LLMMSE) estimator. The entropy/alpha/anisotropy decomposition is then applied to the estimated polarimetric coherency matrices, and coherence optimization is performed on the estimated polarimetric and interferometric coherency matrices. Using this decomposition, unsupervised classification for land applications by an iterative algorithm based on a complex Wishart density function is also applied. The method has been tested on airborne high-resolution polarimetric interferometric synthetic aperture radar (POL-InSAR) images (Oberpfaffenhofen area-German Space Agency). For comparison purposes, the two estimation techniques (complex multilooking and LLMMSE) were tested using three different spatial supports: a fix-sized symmetric neighborhood (boxcar filter), directional nonsymmetric windows, and the proposed AN. Subjective and objective performance analysis, including coherence edge detection, receiver operating characteristics plots, and bias reduction tables, recommends the proposed algorithm as an effective POL-InSAR postprocessing technique.

International audience

This work presents a new method for filtering the coherence map issued from Synthetic Aperture Radar (SAR) polarimetric interferometric data. For each pixel of the interferogram, an adaptive neighborhood is determined by a region growing technique driven by the amplitude image information. Then, pixels in the derived adaptive neighborhood are complex averaged to yield the filtered value of the coherence, after performing a phase compensation step. The proposed method has been applied on airborne high-resolution polarimetric interferometric SAR images. Both subjective and objective performance analysis, including coherence edge detection, shows that the proposed method provides better results than the standard phase-compensated fixed multi-look filter and a linear adaptive coherence filter proposed by Lee et al.

Nonlinear rank-order-based filtering of color images is difficult to implement; the multivariate nature of colors does not allow the introduction of a mathematically-correct and topology-preserving ordering relation. The most widely investigated approach is based on the use of reduced ordering relations introduced according to different scalars. We propose pseudo-morphologic and median operators based on reduced ordering of colors, with respect to scalars computed as geometrical shape invariants of a triangle representation of colors. The same color representation allows the introduction of a luminance invariant intercolor distance, used with good results in distance-based color filters.

Presents a new method for filtering the coherence image issued from an interferometric pair. The basic idea is to determine for each pixel an adaptive neighborhood with respect to the amplitude information. Then, complex averaging is performed using values of pixels in the determined neighborhood to derive the filtered coherence value. It is shown that the proposed technique performs better than the standard fixed-neighborhood filtering technique, both objectively and subjectively.

This contribution proposes a fuzzy approach to color image filtering by the fuzzy modeling of the concept of color credibility. Based on the perceptual notion of color resemblance, the colors are modeled as fuzzy sets in the CIELAB color space. The filtering principle is to select at the filters output the color that is the most credible with respect to the rest of the colors within the filtering window. Although the approach does not make any assumption on the desired filter type, the result is similar to a vector median-type filter.

The aim of the paper is to introduce two new multichannel median type filters, obtained by vector extension of their scalar (gray scale image) counterparts. The two filters are typical examples for two situations in image processing: the use of the fuzzy attribute to justify the weighting factors and the fuzzy attribute as descriptor for rule-based processing.

Vector (multichannel) signal processing has arisen in the last decades as a major area of interest. The classical single-channel (scalar valued) processing techniques cannot deal with the interchannel dependencies, yielding to the development of new methods. This paper investigates two approaches to the median filtering of vector valued signals: filtering based on the reduced ordering in the space of signal samples and filtering with channel decorrelation. A novel filter is proposed that uses channel independentization by the central limit theorem. This filter performs better than classical filters in terms of mean square error measures and visual perception.