Research On Image Defogging Enhancement Algorithm Combining Retinex And Dark Channel

Image is an important way for human beings to obtain information.Image quality determines the accuracy of subsequent image interpretation,analysis and recognition.In recent years,fog and haze weather occurs frequently and distributes widely in China,resulting in a large number of outdoor imaging system performance degradation.Therefore,it is necessary to defog the degraded images in foggy weather,so that the related imaging system can work steadily in bad weather.In this dissertation,image defogging is studied from two aspects: image enhancement and image restoration.In the aspect of image enhancement,the defogging algorithm based on Retinex theory is analyzed in detail.To solve the problem of long operation time and poor defogging effect,an improved frequency domain SSR algorithm based on fusion histogram equalization is proposed,which combines processing domain conversion and histogram equalization algorithm.In the aspect of image restoration,the dark channel prior algorithm is analyzed in detail.Aiming at the problem of time-consuming operation,guided filtering is used instead of soft matting to refine the medium transmission rate and image scaling mechanism is introduced to reduce the time-consuming of the algorithm at the cost of losing the image precision appropriately.In order to improve the applicability of the algorithm,the dark channel image specification and tolerance mechanism are introduced to deal with the color spots in bright areas after processing.Finally,the improved dark channel priori algorithm based on guided filtering and bright region optimization is fused with the frequency domain SSR algorithm based on HSI color model,and the simulation experiments are carried out on the related algorithms in this paper.The experimental results show that the improved algorithm proposed in this dissertation can improve the performance of the original algorithm from different angles compared with the original algorithm,and has a certain application value in the field of image defogging.