Abstract
!edge #$%&' ()*+ ,-. /01234 &56 7. (89:**(; ?@ABC-DE7FGHEI J%&K>LKM ()*+ $%&' #7. #NOP56Q&RDark Channel Prior(DCP) GHEI*9S(T* !U. !>VWX$YZ&[< $%&' 47. &\ARGB ]^ !_`aVb*X cH8.$ d. 7ef 234&gh7. ,-!& GHEI J%&'23h ij$klmn/ opq!KYGHEIrs&'tj$uvwxgyzmH (h{r#:|7.AbstractIn this paper, we propose a transmission rate compensation method to remove a haze of an image by using edge information of a haze image and image segmentation. With a hazed image, it is difficult not only to recognize objects in the image but also to use an image processing method. One of the famous defogging algorithm named 'Dark Channel Prior'(DCP) is used to predict fog transmission rate using dark area of an image, and eliminates fog from the image. But there is a big possibility to calculate a wrong transmission rate if the area of high RGB values is larger than the area of the reference area. Therefore we eliminate color distortion area to calculate transmission rate by using the propose method, and obtain a natural clean image from a hazed image. Keyword : haze removal, fog, haze, dark channel prior
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
