Review of Deep Learning Based Image Dehazing for Autonomous Vehicle

Main Article Content

Pankaj Vinayak Deshmukh , Alok Kumar , Prasun Chakrabarti

Abstract

Computer vision technology plays an important role in distinguishing the surroundings by autonomous vehicles under adverse weather conditions. The images captured under rainy, foggy, or hazy conditions have reduced contrast and color variation due to the scattering of light. Therefore, deep learning (DL) based techniques are used in image dehazing to detect and remove the haze effect in images without relying on any other models. This paper comprehensively reviews the deep learning-based image dehazing methods used in autonomous vehicles. First discussed the background details of image dehazing and then investigated the DL methods used in image dehazing. Reviewed the benchmark datasets used, and performance matrices evaluated in image dehazing applications and then surveyed the applications of DL in image dehazing. Finally discussed are the potential challenges, significant findings, and future research directions required for further study.

Article Details

Section
Articles