Wireless image and video soft transmission via perception-inspired power distortion optimization

摘要: Recently, a scheme called SoftCast has shown great potential for wireless image/video communication in the scenarios where channel quality may fluctuate drastically and unpredictably. The transmission is lossy nature, with its power allocated among coefficients unequally to minimize distortion. One problem that performance optimized using mean square errors (MSE) as metric, which known not matching perception of human eyes. Inspired by researches image assessment, this paper proposes allocation optimization minimizes perceptual distortion reconstruction image. In particular, we establish model evaluate importance different transform coefficients, based on structure similarity (SSIM) metric. Experimental results show proposed can improve original scheme.