Multi - Environment Image Dehazing Algorithm Based on Color Channel Attenuation Difference

In this paper, we propose a multi-environment image dehazing algorithm based on the color channel attenuation difference(ME-DCAD). The algorithm possesses several unique features: firstly, it uses a combination of the Beer-Lambert law and the light square attenuation law to more accurately define the relationship between light attenuation and distance; secondly, it estimates the transmission rate by calculating the difference in attenuation with distance through different color channels, which results in good dehazing effects in both underwater and hazy environments; thirdly, it considers the problem of dust accumulation on the lens and can be applied in environments including scenarios involving coke inspection in factories. Through comparative experiments with similar methods, our proposed method exhibits certain image deblurring and enhancement performance and performs well in various environments, improving the image quality obtained in hazy environments.