MCPA: A Fast Single Image Haze Removal Method Based on the Minimum Channel and Patchless Approach

The image haze removal algorithm is challenging regarding computational processing speed and the hazy effect. Instead of using local patch approach, which assumes scene transmission to be locally constant uses various filters smooth map, this paper proposes a fast single method based on minimum channel patchless approach. A new simple approach estimate atmospheric light proposed images. histogram used extract pixels exclude non-hazy bright in image. equalization multiplication are applied achieve better visual quality. In order verify performance method, 100 images collected from datasets I-HAZE, O-HAZE, websites. Experimental results show that our outperforms up-to-date state-of-the-art algorithms quantitative evaluations. From subjective comparisons, most current color restoration. Also, time assessment fastest among methods about 15 times faster than second-fastest method. main contribution significantly reducing computation because it does not need any filter complicated algorithms. addition speed, can