Quality enhancement for hevc compressed videos

Abstract

The most frequent form of internet traffic is video. Video compression is essential for offering high-quality image/video services. Video coding is one of the main technologies utilized in video applications. High Efficiency Video Coding (HEVC) enables 50% compression while preserving comparable perceived image quality compared to the previous video coding standard H.264/AVC. Complex artifacts are introduced by high compression rates and enhancing HEVC compressed videos is crucial. In-Loop Filtering (ILP) is a technique used by HEVC to greatly minimize artifacts and enhance the quality of the reconstructed video. Deblocking Filter (DBF) and Sample Adaptive Offset (SAO) filters are used in HEVC to minimize artifacts. It is possible to improve the visual quality of decoded images by using deep learning techniques. The quality of HEVC decoded video can be improved using a variety of models, including Deep Convolutional Neural Networks (DCNN), Generative Adversarial Networks (GAN), and U-Net architecture. Among the many applications of deep learning techniques for video quality enhancement, the detection of moving objects is one of the most important. newline