Video Restoration Based on Linear Transformation Method For Real Time Fog –Free Vision In Vehicles

Abstract

The detecting problem of day time fog and calculating visibility distances is the main focus of this paper. An original linear transformation method has been developed for achieving the goal. The solution mentioned here, features the benefit of using a single video camera output. From the camera output the sky region and the fog infected region are identified. The fog infected region is further applied quad-tree subdivision, based on this, an additional method for linear channel is introduced and the calculation of atmospheric light is carried out. Comparing with the other method the linear transformation method requires less time consumption and provides high accuracy. Initially the transformation of the foggy part of the video to the gray scale and gradient scale is taken place then, from the transformed gray and gradient video fragments the transmission map is estimated and further assessment criteria are preceded. By comparing with the previously used algorithms for video processing, the algorithm introduced here is faster. The introduced video processing needs very less requirements for the implementation in vehicles.

Keywords: Fog, image enhancement, video analysis, video processing, video reconstruction, video restoration, weather conditions

Cite this Article
Arunima A.R, Vishnu V.S. Video Restoration Based on Linear Transformation Method for Real-Time Fog: Free Vision in Vehicles. Journal of Operating Systems Development & Trends. 2018; 5(1): 34–40p.