1995, And coding devices and of brief of images, the computer program using them and the support of informaton making it possible to implement them

摘要

The method of coding of a digital image comprises, for at least one macroblock, called the "current " of the image: - a step (225) of the estimation of movement applied to the current macro-block and at least an image called the" reference " in order to obtain at least one first predictor in each reference image, - a step (215) for obtaining a second predictor of a level of spatial scalability lower than the level of spatial scalability of the current macro-block, - a step (230), from the spatial domain to the frequency domain, of at least a predictor representative of at least one first predictor, - a step (235) of the prediction of at least a low coefficient of the spatial frequency as a function of the second predictor and - a step (235) of the prediction of at least one coefficient of a high spatial frequency as a function of at least one coefficient of high frequency of at least a predictor representative of at least one first predictor converted into the frequency domain.