A Study On Image Dehazing Algorithm Based On Fusion Dark-Light Channel Prior

Digital image processing technology has been widely used in medical,security,military,education and other fields,bringing a lot of convenience to our lives.However,due to the influence of haze and other bad weather,there are many problems in collected images,such as low contrast,decreased clarity and image color distortion,which seriously affects the subsequent image processing in recent years.Therefore,how to improve the recovery effect of low quality images under foggy condition is particularly important.At present,the dehazing algorithm based on dark channel priori has achieved ideal effect,however,the recovered image still has the following problems: it is not applicable to large-area bright white regions,the estimation of atmospheric light value is inaccurate and dehazing weight parameters are relatively single.In order to solve the above problems,relevant improvements are made in this paper,and relatively good experimental results are obtained.The main work and innovation points of this paper are as follows:(1)Aiming at the problem that the segmentation of image background and target is not accurate due to the single threshold parameter in image segmentation,this paper proposes a local threshold segmentation method to effectively solve this problem.(2)Against the problem that the dark channel prior dehazing theory is not applicable to the bright white and sky region,this paper proposes a dehazing algorithm combining the dark channel and the bright channel,and proposes the bright channel prior theory according to the feature that the pixel value of the bright white region is generally high.By combining light and dark channels,the improved transmission rate formula is obtained,which effectively solves the problem of large bright white area in foggy image.(3)Aiming at the problem of inaccurate estimation of atmospheric light value,this paper proposes an algorithm to obtain the atmospheric light value by using grayscale open operation and combining with interval estimation weighting,which makes the obtained atmospheric light value more robust.(4)The weight that affects the degree of defogging is relatively single in the existing defogging algorithms,a self-adjusting weight optimization defogging algorithm is proposed in this paper to solve this problem.Through a large number of experiments,it is found that there is a certain relationship between the defogging weight and the average pixel value of the image,based on this relationship,the algorithm proposed in this paper is improved to make the image more real and natural after defogging.After proposing the improved algorithm in this paper,a variety of defogging algorithms are selected and compared with the algorithm in this paper,and a subjective and objective evaluation method is adopted.Through subjective analysis and objective data,it is proved that the algorithm in this paper effectively solves the problem that the original algorithm is not applicable to large-area bright white area.Furthermore,the visual effect of the restored image based on the algorithm in this paper is more real and natural,closer to the actual scene.