Home About Login Current Archives Announcements Editorial Board
Submit Now For Authors Call for Submissions Statistics Contact
Home > Archives > Volume 20, No 8 (2022) > Article

DOI: 10.14704/nq.2022.20.8.NQ44084

NEURAL NETWORK BASED CONCEPTUALIZATION AND DESIGNOF A TWO-STAGE VIDEO COMPRESSION ARTEFACT REDUCTION STRATEGY

Dr.C.RAVICHANDRAN, Dr.T.THIRUMURUGAN, Dr.C.SENTHAMARAI, Dr.C.KALAISELVAN

Abstract

Multimedia communication has had a huge impact on data transmission and reception in recent years. Video transmission, reception, and storage are all important aspects of multimedia communication. The requirement to store the movie on the drive necessitates a significant amount of space and memory. Lossy compression and lossless compression are the two types of video compression techniques. Lossy compression compresses video frames more efficiently than lossless compression, but the video quality suffers as a result of artefacts. When compressing the video frame, an artefact reduction technique is employed in this study. Despite the fact that video frame data compression techniques have the ability to dramatically reduce the quantity of video frame material, they also generate visual distortions due to lossy compression. The proposed method increases video quality after compression by employing the DREPF technology to remove artefacts. DREPF has been proposed as a way to improve decoded video frames after compression using a technique called Recursive Ensemble Particle Filtering (REPF). Recursive Ensemble Particle Filtering (REPF) locations are determined using the DCNN, which is utilized to include them into the deep Recursive Ensemble Particle Filtering (REPF). For better video frame enhancement, the previous information is approximated by combining the time and temporal systems. Model-based and learning-based systems can be combined in a novel way by combining the Recursive Ensemble Particle model's recursive structure with DCNN's dominating illustration capabilities. Experiments have shown that the proposed methodology produces better outcomes than current systems.

Keywords

Deep Convolutional Neural Network (DCNN), Deep Recursive Ensemble Particle Filter (DREPF), Lossless compression, Lucy–Richardson (LR) algorithm

Full Text

PDF

References

?>