Research Of Defog Algorithm Based On Partial Differential Equation And Image Enhancement

Under the foggy conditions,there are a large number of aerodynamic diameter particles less than 2.5 microns in the atmosphere,which can refract and reflect large amounts of visible lights,causing blurred images taken by the image device.Therefore,it is very important to discuss the image defog algorithm.Firstly,this paper analyzes the causes of the degradation of the fog images,and summarizes the classical defogging algorithms based on image enhancement and image restoration both at home and abroad.From the perspective of image enhancement,the original algorithm is improved for the image defog algorithm based on the distinction of sky coverage area.Firstly,by converting the image color space,the image luminance component is extracted,and the color component and saturation component are kept stable.The brightness component is segmented according to the preset numbers of clusters.The pixels with the largest and next largest number of clusters are selected as the sky region.The brightness of the image sky region is kept stable,and the non-sky part is processed by contrasted limited adaptive histogram equalization.The final result is obtained by using the image fusion technology to deal with the processing of the sky coverage area and the rest area.Experimental results show that the details of the image are clearly seen and have a good subjective feeling.The speed of the algorithm is much higher than that of the original algorithm.Secondly,from the perspective of image restoration,a new full-variational image defog algorithm is proposed to improve the image darkness and partial color distortion for the full variational image defog algorithm.On the basis of the original algorithm,the image vanishing point is selected by interactive way;and the normalized distance between the image vanishing point and the remaining pixels is calculated.According to the dark channel map,the pixel point of 0.1%before the dark channel value is obtained,and find out the dark channel value in the corresponding position of the original image.The sky brightness value is calculated by the average of the 3x3 pixels in the neighborhood of those pixels.The total variation method is used to solve the corresponding partial differential equation.The defog recovery image continues to be subjected to secondary image enhancement using the contrast limited adaptive histogram equalization.The experimental results show that the improved variation image defogging algorithm not only achieves good results in subjective feelings,but also more obvious details of the image,and can keep the original color information of the image,and the calculation speed of the algorithm has a certain degree of improvement.