An Improved Dehazing Algorithm Combining Dark Channel Prior and Artificial Bee Colony Algorithm
Zhenguo Li, Lisheng Ma, Yibin Hou, Jiaxin Li, Tao Tao · 2024
Aiming to address significant issues like severe halo effects and excessive noise present in images processed by the traditional dark channel prior-based dehazing algorithm using fixed values, a proposed enhancement method leverages the characteristics of the artificial bee colony algorithm for parameter optimization. It seeks the optimal atmospheric light transmission factor values within different average image brightness ranges and integrates these values into the dark channel prior-based dehazing algorithm. Simultaneously, in estimating the atmospheric light, a median filtering algorithm replaces the original twostep minimum value filtering process. This improvement exhibits notable effectiveness in reducing halo effects and noise compared to the traditional dark channel prior-based dehazing algorithm.