Abstract
AbstractThe paper proposes an efficient method of removing haze from outdoor images using dark channel prior (DCP) method. The DCP represents the statistics of outdoor images. In dark channel images, minimum of one-color channel of haze-free images from outdoor locations has pixels having very low intensity. Therefore, using imaging model of dark channel prior (DCP), we can identify the quantity of haze in an image. Moreover, the transmission map estimation and refinement using guided filter are done after the application of DCP which improves the transmission map of the image. The comparative values of PSNR and RMSE are identified for images of the transmission map obtained from DCP image and are found to have minimum value for the guided filter. The results based on the proposed method are demonstrated for generating DCP, transmission map estimation, refinement, and recovery of radiance of images. It is imperative to say that this method succeeded in its aim to transform haze images into haze-free images with high-intensity pixel values.KeywordsDCPDark channel priorTransmission mapEstimationScene radiance
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
