Colour balance and contrast stretching for sand‐dust image enhancement

The increasingly frequent sand-dust weather in the inland areas seriously affects outdoor vision applications, especially autonomous vehicles and security monitoring. To moderate image's colour cast poor contrast caused by weather, an effective approach is proposed this study to enhance images. First, original degraded corrected a new balance compensation formula, which compensates blue green channel information through numerous yellow scattering before white balance. Next, order avoid deviation, image converted from RGB space HSV use CLAHE V component improve contrast. Then, nonlinear gain function defined further adaptively sharpen details. Finally, S stretched saturation. extensive qualitative quantitative evaluation shows that method can edge clarity contrast, restore good fidelity for all images tested. verification also proves of much significance improving feature point extraction target detection results weather.