A Novel Visibility Restoration Technique for Hazy Videos Captured in Real - World Weather Conditions

The visibility of outdoor videos were seriously degraded under hazy, foggy and sandstorm weather conditions, which will affect the computer vision applications, such as outdoor object recognition systems, obstacle detection systems, video surveillance systems, and intelligent transportation systems. This paper proposes a novel method for restoring the visibility of degraded videos that uses a combination of three major modules 1) a depth estimation (DE) module, 2) a color analysis (CA) module, and 3) a visibility restoration (VR) module. This visibility restoration approach is based on the conjunctive utilization of the median filter and guided joint bilateral filter operation, the adaptive gamma correction technique, the gray world assumption, and the dark channel prior method. The proposed DE module takes advantage of the median filter technique and adopts adaptive gamma correction technique. By doing so, halo effects can be avoided in videos with complex structures, and can achieve an effective estimate of the transmission map. The proposed CA module is based on the gray world assumption and analyzes the color features of each input hazy frames. Subsequently, the VR module uses the adjusted transmission map and the color-correlated information to mend the color distortion in variable scenes captured during inclement weather conditions. The proposed method will successfully be applied to videos, hence we can use this algorithm in even real time video applications.