a60图像压缩变换学习.pdfVIP

Graph Transform Learning for Image Compression Giulia Fracastoro Dorina Thanou, Pascal Frossard Department of Electronics and ecommunications Signal Processing Laboratory (LTS4) ´ Politecnico di Torino E cole Polytechnique Fe´de´rale de Lausanne —In this paper, we propose a new graph-based com- straightforward, the problem of designing a graph transform pression scheme for image coding. Our approach relies on the stays critical and may actually represent the major obstacle careful design of a graph that optimizes the overall rate-distortion towards effective compression of images. performance. In particular, we model the pixels as nodes of a graph and we treat the pixel intensities as a signal living on In this work, we propose a novel graph-based framework an unknown graph topology. We then introduce a novel graph for effective coding of images that takes into account the learning algorithm targeted for image compression that uncovers coding of the images as well as the cost of transmitting the the connectivities between the pixels, by taking into consideration graph. In particular, we introduce an innovative way for coding the coding of the image signal and the graph topology in rate- the graph by treating its edge weights as a graph signal that distortion terms. The cost of the graph description is introduced in the optimization problem by treating the edge weights as lie