Archives

Video Enhancement Technique Based on Horn Schunck Optical Flow Estimation and Laplacian Pyramid Segmentation


Padma Reddy A M,Udaya Rani V,
Abstract

The video processing techniques may introduce some quantity of artifacts or distortion in the video signal. The video enhancement techniques are used to improve the system performance in terms of coding efficiency, error resiliency, capability of video analysis, and visualization quality. Nowadays, enhancement of video quality is one of the significant challenges in video processing. However, videos are frequently affected by noise during transmission, acquisition and processing. The noise destroys the video structure and reduces video quality. So, Horn Schunck - Laplacian Pyramid technique is used to improve the video quality. In this research, the optical data are computed by using Horn Schunck technique to compensate the blurring of moving image. The Laplacian Pyramid allows a smooth enhancement of the large video structure, which avoids the visible artifacts. The Laplician and median filter output is given to the fusion block as an input. The motion compensated image filtered by a hybrid median filter and those filtered values are fused with the chrominance and Laplacian component. Finally, enhanced video quality is obtained. The Horn Schunck-Laplacian Pyramid technique is implemented in the MATLAB version 2018 software tool. The experimental results showed that the Horn Schunck-Laplacian Pyramid technology improved the quality of the video up to 5-10% compared to the existing methods.

Volume 11 | 02-Special Issue

Pages: 1802-1811