UAV target detection based on improved dark channel prior defogging
Image fog removal is an important research focus in image processing field.In order to solve the prob-lem of image defogging and enhancement in hazy weather,a defogging algorithm based on improved dark channel is proposed.First of all,in order to make the haze image closer to the non-fog image and improve the clarity of the im-age,the algorithm reduces the RGB channel values of the fog image respectively,and combines each reduced channel and the other two previously unreduced channels,and then uses the image de-fog algorithm to weight three new images to restore the image.In order to solve the problem of color distortion in the sky region of the image,a parameter K is set to calculate the transmittance of the sky region and the non-sky region respectively.In order to solve the problem of over-dark brightness and increase target contrast,this paper introduces CLAHE method to enhance image process-ing.The experimental results show that:The contrast value of the proposed algorithm in the five images is more than twice and more than three times that of the MDCP and DCP algorithms respectively,and the average information entro-py in the five images is 7.558 9,which is obviously better than the other two algorithms.Moreover,the average accu-racy of the proposed algorithm in target detection under haze weather can reach 73％,which is 7％higher than before the improvement,and has certain feasibility.
fog removal enhancementdark channel modelcolor channeladaptive skyCLAHEuav view tar-get detection
NETL
NSTL
万方数据
