A universal video enhancement technique based on least path intensity detector and adaptive filtering algorithm

Noise removal plays an important pre-processing stage for motion estimation and video enhancement. This paper describes a novel algorithm for color video sequence enhancement using Least Path Impulse Detector extended to spatio-temporal filtering. This algorithm works well with both Gaussian and impulse noise and quality analysis factor such as PSNR and SSIM proves the efficiency of proposed method. The simulation of available algorithm for mixed noise removal was implemented and compared with proposed method and results are tabulated. The algorithm can be further extended for enhancement of highly compressed video sequences.