The paper presents a watermarking technique, specific to color images. The insertion and the detection are based on the 2D discrete wavelet transform, applied on each color component. Three color vectors are extracted from the wavelet coefficients. The insertion consists of modifying one vector for each location, with regard to the bit value and the vector triplet. The mark is extracted without the original image, by observing the scheme of each vector triplet. This new method has been shown to be resistant to JPEG compression, median filtering and noise adding. Moreover, each insertion is weighted by an adaptive strength, obtained by a retroactive process between the original and the watermarked images. This process allows optimizing the compromise between invisibility and robustness, considering the local image color content.

We consider a new way to encode video sequences. The proposed method is based on second generation wavelets (SOW), a novel mathematic transform, successfully applied in 3D coding. This video coder is created by combining the wavelet theory and mesh geometry. With this and the SOW, we will be able to apply onto the video image, the mathematical transform to locate and to encode the error, results of the motion compensation process. This powerful signal processing theory allows a very well adapted coding of high frequencies. The main advantage is that these wavelets can be designed to fit exactly on singularities, shapes, textures, and edges. Hence, we can reduce redundancy and improve considerably coding efficiency over those peculiar settings. This new coder can be applied on all classical applications, such as very low bit rate transmissions, video streaming, vision conference and video on demand.

This paper presents a video coding scheme combining H264 with a triangular mesh based representation. In this scheme, we developed a mesh-based motion estimator in which the motion vectors are computed at each node of the mesh. All other motion vectors are then interpolated using the nodes denoting the patches. Finally, this model is integrated in the RD optimized model of the H264 coder. To end, we propose some results, which compare H264 with our hybrid approach. This approach permits us to improve the quality of P and B images obtained by H264 for the same coding cost.

Detecting human skin in color images is an interesting problem due to the potential number of target applications ranging from face detection to the detection of pornographic images. Many approaches have been proposed in literature including color thresholding, statistical skin color modeling and segmentation-based techniques. However, the performances often decrease in the presence of illumination artifacts resulting in sub-optimal results. In this paper, we propose a new method that is adaptive to the image content especially in illumination conditions. For this purpose, we first detect skin seeds that serve as the basis to locate the skin colors with highest density. This step allows to take into account illumination conditions for the last step, namely the growing of skin seeds.

This paper proposes a new video watermarking scheme based on a hierarchical motion analysis, that is robust against classical video processing (filtering, lossy compression...). This model works in the uncompressed domain and disturbs the motion vectors computed by an exhaustive BMA on blocks of size n*n in order to insert an invisible watermark enabling a protection against most attacks. Finally, to increase the robustness, our system inserts the mark by generating a hierarchy of motion vectors of size N*N with N=k*n to spread the mark on the lower level associated to block n*n.

Multimedia data security is very important for multimedia commerce on the Internet such as "pay-per-view" services. Thus, watermarking algorithms for data security appear. These algorithms describe methods and technologies that allow information to be hidden, for example a number or text, into a media, such as images, video, audio files... In this paper, we propose a new application of watermarking based on a scrambling process. In this application, we develop a waterscrambling technique in which video data are scrambled efficiently by disturbing a subset of motion vectors. The interest of this approach is its ability to scramble a video while maintaining a certain visibility. Effectively it permits the use of considerable levels of security in order to choose the level of perceptibility of the video. By using watermarking techniques we will be able to combine this approach with a classical copyright watermarking system.

A new very low bit rate scalable encoding principle is presented. It is based on hierarchical triangular mesh encoding combined with triangle based DCT coding. Affine or DCT approximation is selected according to best rate distortion trade-off. In this way our coder demonstrates advantages at very low bit rates avoiding the traditional block or ringing artefacts of current DCT or wavelet based coders.

We present a method which is able to determine the occlusion areas appearing in a mesh-based motion estimation. The use of the motion estimation scheme based on a non-uniform hierarchical triangular mesh and the optimization of the nodal motion vectors by means of a multi-resolution differential method leads to efficient and fast motion estimation as well as motion field refinement. Although global motion estimation avoids the blocking artifacts due to the independence of the neighboring blocks of conventional block matching, it can only find the optimal motion parameters, which derive from the motion vectors at the node points that are shared by several patches of a continuous field. However real motion in a video sequence is discontinuous. Thus, our goal consists of modeling the artifacts, resulting from the occlusions by using a global mesh. Contrary to the segmented mesh associated with the video object when fast motion has to be estimated, surface mesh may contain overlapping triangles and the estimated image is thus deteriorated. In fact, these overlaps allow the detection of the occlusions. Thus, our method consists of taking into account this information to determine the optimal motion vectors in the continuous field and to generate the content of the uncovered area or to exclude the content of the covered area. To illustrate the new performances of the proposed method, results showing the capabilities of our mesh-based motion estimation to determine discontinuities lines are presented.

This paper presents a video coding scheme using a triangular mesh based representation. In our approach, the video content is supported by a surface model based on the finite elements technique, offering a high flexibility and enabling numerous processing and functionalities. To complete this scheme, we developed a mesh based motion estimator in which the motion vectors are computed at each node of the mesh. All other motion vectors are then interpolated starting from the nodes denoting the patches. Finally, the obtained motion vectors are applied to the intra model representation, performing thus forward mesh tracking along the video sequence. To end, we propose some results, which compare MPEG4 with our approach.

In this paper, we present an evolution of IFS-based image compression schemes replacing usual affine mass map by cosine mass map, well adapted to high frequency contents and poor domain pools. This change implies the creation of a new IFS determination algorithm for solving the inverse problem, and a new way to quantize and store involved coefficients. We then present how VOP facilities can be performed using this new IFS family.

We introduce a mesh approximation method that uses a volume-based metric. After a geometric simplification, we minimize the volume between the simplified mesh and the original mesh using a gradient-based optimization algorithm and a finite-element interpolation model implicitly defined on meshes. The notable contribution of this paper is the theoretical framework which permits the construction of a volume minimization process between two triangular meshes. We chose this volume-based metric because of its good perceptual properties, as it naturally and accurately fits the geometric singularities on 3D meshes. Furthermore, this metric corresponds well to a sort of intuitive error between two 3D surfaces and the resulting optimization algorithm only requires a few parameters. We show that this approach permits geometric compression leading to multiresolution meshes with minimal visual losses.

We present a method for avoiding triangles overlapping in mesh-based motion estimation using an augmented Lagrangian function. The use of the motion estimation scheme based on a non-uniform hierarchical triangular mesh and the optimization of the nodal motion vectors by means of a multi-resolution differential method leads to efficient and fast motion estimation as well as motion field refinement. However, when high motion has to be estimated, the surface mesh can contain some non-justified triangles overlapping and the estimated image is thus deteriorated. To avoid this situation, an augmented Lagrangian function is added to the prediction error. To illustrate the new performance of the developed method, results comparing this method with the classical one concerning the motion mesh and the estimated image are presented.