Shape-adaptive model-based codec for lossy and lossless compression of images

摘要

The present invention relates to methods and codecs for image compression and video compression. Embodiments of the present invention support binary shape and matte segmentation and soft segmentation image compression by decomposing the input shape for flexible lossy and lossless coding into deterministic and probabilistic components A new shape adaptive model-based codec (SAM). The present invention can provide inter / intra prediction and flexibly fits between lossy and lossless modes with various parameters for compression quality control. The compression module can also be adapted using a number of other compression techniques. [Selection] Figure 3