基于暗通道的沙尘图像增强算法

针对沙尘图像存在色差以及图像增强后沙尘图像在明亮区域易出现光晕等问题, 提出一种新的基于暗通道的沙尘图像增强算法. 首先通过在Lab色彩空间用灰度世界算法调整色差, 有效避免图像出现色彩失真现象; 然后利用伽马校正函数和暗通道去雾算法, 避免图像出现噪声、 色彩过度增强和光晕等现象; 最后将亮度补偿后的图像与对比度增强后的图像进行加权融合, 进一步提高图像的可见度, 使图像细节更清晰可见. 实验结果表明, 该方法提高了沙尘图像的清晰度和亮度, 可得到质量更高的沙尘图像.

Dust Image Enhancement Algorithm Based on Dark Channel

Aiming at the problems of chromatic aberration of dust images and the tendency of halo of dust image in bright areas after image enhancement, we proposed a new dust image enhancement algorithm based on dark channel. Firstly, by using grayscale world algorithm to adjust chromatic aberratin in Lab color space, the phenomenon of color distortion could be effectively avoided. Secondly, gamma correction function and dark channel defogging algorithm were used to avoid image noise, excessive color enhancement and halo. Finally, the brightness compensated image and the contrast enhanced image were weighted and fused to further improve the visibility of the image and make image details more visible. The experimental results show that the method improves the clarity and brightness of dust image, and can get higher quality dust image.