Study on the Application Domain of Two Traditional Defogging Algorithms Based on Atmospheric degradation model

Abstract

The development of artificial machine vision enables further research on image processing. As the direction of image enhancement, research on image defogging is also essential. Base on Atmospheric degradation model, the image defogging algorithm can execute precisely. Dark Channel Prior algorithm (DCP) and Color Attenuation Prior algorithm (CAP) are two classic algorithms in the traditional defogging algorithm. These two algorithms were implemented based on the degraded model and carried out a comparative analysis. By comparison, it was found that DCP performs better on images with no sky in the middle and short distances, while CAP performed better on images with larger sky areas and deeper depth of field. Regarding the difference between the two algorithms, further research was performed and the reason for the difference was attributed to the calculation method of transmittance, and the reasons for the difference between the two defogging algorithms were summarized.