改进暗通道先验的低光照缝隙图像增强算法

Research on Low Lighting Gap Image Enhancement Algorithm Using Improved Dark Channel Prior

针对工业环境中缝隙图像由于低光照而产生的模糊不清和对比度低等问题,提出一种基于改进暗通道的缝隙图像增强算法.首先将输入图像进行滤波并反转;然后采用改进的暗通道估算大气光值和初始透射率,利用引导滤波和中值滤波优化初始透射率,达到平滑图像和保持边缘的目的;最后将增强后的反转图像反转,实现低光照缝隙图像增强.实验结果表明,该算法能够有效地改善低光照缝隙图像的感知质量.

The gap images taken under industrial environment usually have serious loss of visibility and contrast due to low light.To solve the problem,an enhancement algorithm of gap image is proposed based on improved dark channel prior.The algorithm worked by first filtering and inverting an input image and then applied an improved dark channel to estimate the atmospheric light and initial transmission.The guided filter and median filter were used to optimize the initial transmission for smoothing the image and maintaining the edge.The final enhanced image was obtained by the inversion of the enhanced inversion.The experimental results demonstrate that the proposed approach can effectively improve the perceptual quality of low light gap image.