Research On Digital Image Haze Removal Algorithm Using Dark Channel Prior

At present,based on the rapid development of image processing technology,computer vision systems are becoming more and more popular in life.Most computer vision systems need to first extract the feature features of the acquired image before performing image recognition,analysis,etc.,and the accuracy of feature extraction requires that the acquired optical image has a certain definition.However,images acquired by outdoor imaging devices are highly susceptible to weather.In the case of fog,the light is affected by the scattering particles in the fog,so that the light received by the optical sensor is affected,and the quality of the formed digital image is reduced,which is mainly caused by the decrease of local contrast and the blurring of the scene,making the image details difficult.Get,while looking white,the color saturation drops.This directly affects the work of outdoor imaging equipment,bringing huge security risks to people's production and life.Therefore,it is of great significance to study how to improve the restoration effect of foggy degraded images and reduce the adverse effects of haze weather on outdoor imaging systems.After analyzing the causes of degradation and degradation of foggy images and grasping the current image dehazing algorithms,this paper deeply studies the dark channel prior image dehazing algorithm proposed by He Yuming and so on,and analyzes the defogging of this algorithm.The principle and the inadequacies that need to be improved: First,when the algorithm deals with foggy images with large-area sky areas,the restored images often have obvious color distortion,and the halo phenomenon is prone to occur in scene mutations;It is the overall darkness of the processed fog-free image,and the color contrast is not good.Thirdly,it is too long to use the soft-strip method or the guide filter to estimate the transmittance of the system,which cannot meet the real-time requirements of the system.Aiming at the above problems,this paper proposes a fusion-based image dehazing algorithm.On the basis of the dark channel prior,the original single is selected by selecting different sizes of filtering windows,and then the double-tree complex wavelet transform is used to fuse the minimum color channel and the dark channel of the atomized image,and the quadtree subspace is used for hierarchical search.The method optimizes the estimation of atmospheric light value,and obtains a finer transmission map,and then performs image enhancement processing on the defogged image.Finally,the fog-free image processed by the proposed algorithm and various classical image dehazing algorithms are adopted.Experiment simulation was performed on the same platform,and the simulation results were subjectively and objectively evaluated.The simulation comparison experiment proves that the improved algorithm proposed in this paper has better defogging effect on foggy images,the halo phenomenon is eliminated,the sky area color is natural,the image brightness recovery is moderate,and the algorithm real-time performance is improved.