A computational framework to Investigate the performance of Deraining algorithm for hevc video Encoder

Abstract

High-efficiency video coding (HEVC) standard is the advanced and widely used video coding standard in recent High Definition(HD) video applications. This video codec faces many challenges during video transmission over the lower bandwidth channel. One of the main challenges is that rain streaks removal from HD video sequences. The video capturing devices, such as digital cameras and smartphones are popular nowadays and inexpensive. It is used to capture high-quality video sequences in any environmental condition. But the quality of these video sequences is despoiled due to various outdoor weather conditions such as raindrop, snow, haze, and fog. Among the four outdoor weather conditions, rain streaks consent the complex pixel variation in captured outdoor video sequences. This rain streaks affect the background scene of video sequences and introduce the temporal misalignment in consecutive video frames. During ransmission of such outdoor video sequences corrupted by rain patterns using HEVC encoder, the coding performance of encoder is mainly exaggerated. Rain streaks removal is an important pre-processing step in the real-time image and video application. For noise removal, a video encoder has in-loop filters specifically. Although these loop filters are efficient, they are not effective for removing rain pattern from the video sequences. To detect and remove the different intensity level and directional oriented rain patterns three different deraining approaches have been introduced in this thesis. This deraining filter unit is dded with existing in loop filter of HEVC encoder and the performance of encoder is studied for various quantization parameter values. newline