Abstract: The invention carries out a movement compensation of contouring defects. The movement compensation is carried out, for each subfield, by assigning, to each cell, the state which would correspond to a movement-compensating intermediate image located at the instant of said subfield. The method of the invention associates a single movement vector Vm with each cell Ci so as to constitute an intermediate image for each subfield.

Abstract: The invention concerns a device for creating semantic browsing options comprising a means to select an initial set of images from an input database comprising a plurality of images, each image being associated with a semantic information out of a semantic information space, a means to determine a semantic information sub-space represented by the semantic information associated with said initial set of images, The device comprises also: a means to select a reduced sub-space out of said sub-space, a means to create semantic links to images of the database associated with the semantic information of said reduced sub-space, a means to display at least one image of the database associated with the semantic information of said reduced sub-space.

Abstract: The invention relates to a device and a process for detecting faces in a colour image. The process is such that it comprises the steps of: detection of the flesh coloured pixels in the image, segmentation of the image into regions of homogeneous colour, selection, among the regions of homogeneous colour, of the regions comprising a number of flesh coloured pixels greater than a first predetermined threshold, the regions selected being said to be flesh coloured regions. merging of the neighbouring flesh coloured regions until the merge cost attains a second predetermined threshold, the merge cost linearly weighting a merge cost relating to the shape of the regions and a merge cost relating to the colour of the regions, exclusion of false positives.

Abstract: A device and a method for measuring visual similarity between two images. One image (Q) being referred to as the model and one image (T) being referred to as the target, the method comprises a prior step (E2) of segmenting the images into regions (Qi, Ti), with each region there being associated at least one attribute (F) representative of at least one characteristic of the region. It furthermore comprises the steps of calculating (E3) the visual similarity between the pairs (Qi, Ti) of possible regions of the two images (Q, T), by taking into account the distance (D(Qi, Ti)) between the said attributes (F) of the regions (Qi, Ti) matched and the areas of the regions (Qi, Ti) matched; selecting (E4) a certain number of pairs (Qi, Ti) of regions whose similarity is greater than a first fixed threshold (?), calculating (E9) the global similarity between the two images, based on the pairs (Qi, Ti) of regions selected.

Abstract: For identification of tokens in video sequences, first the sequence is scanned for boundaries between consecutive parts of the sequence, where appearance of the tokens is expected. After that, candidate regions are pre-selected and classified in parts of the video sequence adjacent to detected boundaries and tokens are located in the candidate regions. Information carried on the tokens is located, interpreted and merged into consistent sets of information after passing a confidence analysis. In parallel, changes in the visual appearance of the token signaling a special event are detected.

Abstract: The invention concerns a device for generating mutual photometric effects. The invention concerns also a server for delivering photometric parameters for generating mutual photometric effects and a system including such a device and such a server.

Abstract: The invention carries out a movement compensation of contouring defects. The movement compensation is carried out, for each subfield, by assigning, to each cell, the state which would correspond to a movement-compensating intermediate image located at the instant of the said subfield. The method of the invention associates a single movement vector Vm with each cell Ci so as to constitute an intermediate image for each subfield.

Abstract: The invention concerns a method for displaying video images on a plasma display panel. The invention is particularly applicable in plasma display panels comprising a matrix of elementary cells which can be either lighted or unlighted. The invention is characterized in that in order to provide motion compensation of irregular contours, the method consists in defining in the temporal integration window a reference time and in moving the underscans of the temporal integration window relative to said reference time. The underscans consecutive to the time reference in the temporal integration window are moved in the motion direction and the other underscans are moved in the opposite direction.

Abstract: The invention relates to a method for the fragmentation (F) of images (14i) into homogeneous regions (Ri), this fragmentation (F) using iterative merges of fragments Fi and Fj which are as similar as possible according to at least one selection parameter, this similarity being evaluated by a product A*B of two factors A and B, A being consistent with a number of pixels and B being consistent with the selection parameter(s).

Abstract: The generation of images is done by the computation of the motion vectors representing the evolution from one image 2D to the other, directly from the co-ordinates and coefficients of geometrical transformation, of associated points in the 3D scene. The applications relate to the generation of monocular sequences by temporal interpolation and of stereoscopic sequences by extrapolation.

Abstract: A process for estimating disparity between images of a stereoscopic image by carrying out a hierarchical decomposition of each monoscopic image (Ir(t), Il(t)), by levels of resolution in order to build at least three pyramids of images of increasing resolution including one low-frequency image pyramid and two high-frequency image pyramids in the horizontal and vertical directions of the images. Subsequently a disparity D is estimated at the coarsest level of resolution and the estimate is refined in each level as the resolution increases.

Abstract: In this process the frames are composed of frames with parity of a first type (1, 2) interlaced with frames with parity of a second type. It consists in interpolating the missing lines in the direction of the motion (3) on the one hand, by vertical intra-frame linear filtering (6) on the other hand, and in switching between these two values (8) depending on the confidence accorded to the motion vector.

Abstract: The method consists in effecting a hierarchical decomposition of the image by levels of resolutions, (I.sub.31, I.sub.32, I.sub.33 ; I.sub.21, I.sub.22, I.sub.23 ; I.sub.11, I.sub.12, I.sub.13 ; I), in estimating an overall movement at the coarsest level of resolution and in refining the estimate (I.sub.31, I.sub.32, I.sub.33) in each level as the resolution increases.