Research of algorithm for single gray-scale image haze removal
Yu-jiao Shen, Junju Zhang, Si Tian, Kai Zhu, Ying-wang Feng · Optical Sensing and Imaging Technologies and Applications · 2018
The gray-scale image is widely used in remote sensing image and high-resolution image. The resolution and the contrast are declining, also the image quality is seriously damaged under haze weather. Non-model image enhancement mainly uses targeted image processing methods to improve the contrast and details, while the image degradation is considered in model image restoration. Considering the lack and the necessity of research on the gray-scale image haze removal, a method of single gray-scale image haze removal based on dark channel prior is proposed. This paper extends the method of dark channel haze removal to the gray-scale image. The method reduces the amount of calculation by sampling down the input single gray image. At the same time, different measures are taken on the edges, the flat areas and the noise points of the image to remove the block effect. In addition, a refined transmission is obtained by coefficient modulation. By comparing the experiments and the quality evaluation with other methods of haze removal, including guided filtering, gray stretch and adaptive histogram equalization with limited contrast. It fully shows that the proposed method can be effectively applied to the gray image haze removal.