Research On Defogging Algorithm Based On Image Fusion Strateg

In foggy scenes,due to the absorption and scattering of natural light by particles in the air,the image quality taken by the equipment deteriorates seriously,which not only affects the visual perception,but also impeds the subsequent application of traffic monitoring,military reconnaissance and other technologies.Therefore,it is of great significance to study the defogging of images in foggy scenes.The main work completed in this paper is as follows:Aiming at the problems of block effect and high algorithm complexity in dark channel prior defogging algorithm,an improved algorithm based on dark channel prior is proposed.First by improving the minimum filter to estimate the atmosphere light and transmittance,than through the peak signal-to-noise ratio adaptive adjusting parameters to obtain the better row transmittance,roughly the optimized transmittance is obtained by soft matting algorithm,then the row transmittance and the optimized transmittance as input vector of a multilayer perceptron respectively and the target vector for training,By establishing the mapping between row transmittance and optimized transmittance,that the optimized transmittance is obtained instead of soft matting algorithm,and the rough defogging image is restored by combining with atmospheric scattering model.Compared with several classical algorithms,the experimental results show that the improved algorithm can effectively improve the blocking effect and improve the efficiency of defogging.Aiming at the problems of color distortion and less details in the area with high fog concentration in image fusion algorithm,an improved algorithm based on image fusion strategy is proposed.Firstly,the foggy images are white balanced Processing and Gamma correction are performed to obtain the first and second input images of foggy images,and then the rough foggy images obtained by the dark channel prior fogging improved algorithm were introduced as the third input image of this algorithm.The brightness,exposure and significance weight maps corresponding to the three input images are obtained respectively.Finally,the multi-scale pyramid fusion method is used to fuse all input images and weight images,so as to extract their characteristic information to the maximum extent and restore the image to a high quality.Experimental results show that the improved algorithm in this paper can effectively improve the distortion and enhance the detail information in the area with high fog concentration,and has a further improvement in visual effect compared with the dark channel prior dehazing algorithm.