A New Hybrid Post-Filtering Technique for Efficient JPEG Artifacts Reduction

The Joint Photographic Expert Group (JPEG) is the most popular lossy image compression standard owing to its simple structure and economical hardware implementation. However, at the low bitrate compressions, the JPEG standard suffers from the highly visible blocking artifacts, which is the primary reason behind its poor compression performance. In order to efficiently address this issue, this paper forwards a new hybrid post-filtering technique for the JPEG blocking artifacts reduction. The proposed hybrid technique is a proper fusion of iterative steering kernel regression-based filtering and the guided image filtering techniques. Numerous tests have been carried out to validate the blocking artifacts reduction capability of the proposed hybrid technique against the state-of-the-art techniques. The resultant findings confirm that the proposed hybrid technique offers a noteworthy reduction in the JPEG artifacts as compared to the available techniques.