查看文章

In compressive sensing (CS) of images, a block‐based framework is preferred to avoid the huge memory and computation required for a frame‐based approach. However, the recovered image suffers from blocking artefacts due to independent block processing, especially at a low subrate. As a result of this reported work the artifacts are reduced by weighted averaging adopting two techniques: overlapped CS recovery and adaptive weighting. Simulation results show its improvement in both subjective and objective qualities.