Research And Implementation Of Enhancement Algorithm For Low-Illumination Image

In the practical application of computer vision systems, they usually have been used in low illumination, such as night vision or backlight. The images acquired in the situation have low contrast, information lost seriously and color distortion seriously, it is hard to extract the effective information from the image.Therefore, it has a very important practical significance to research the low illumination image enhancement algorithm.Firstly, the paper introduces traditional low illumination image enhancement algorithms, and analyzes their advantages and disadvantages. Then, based on the adaptive adjustment of pupil and visual cell to the environment, the paper improve a new method based on visual perception. Experiments show that the proposed algorithm can effectively improve the performance of details in dark and highlighted area, and it can improve the performance of the image analysis and recognition systems in the low illumination environment effectively, but efficiency needs to be improved. Next, after analysing the relation of low light level images and foggy ones, a low lighting image enhancement algorithm based on physical model was improved. Experimental results show that the proposed algorithm is faster and performs well without losing information comparing with the existing algorithms, meanwhile it can improve the efficiency of image analysis and recognition systems. At last, based on the two kinds of low illumination enhancement algorithms, an image processing software is designed,which has functions of dehaze, low-illumination enhancement, and enhance images captured in water or sand. The software has a very broad application prospect.