An Efficient Edge Detection Technique for Hazy Images using DCP

Main Article Content

Y Saroja, I.sheeba, M. Sudhakar

Abstract

Images of outdoor areas are typically degraded in quality by its turbid medium in the nature such as haze, fog and smoke. The absorption and scattering of light on such kind of images effects the quality of the image. The degraded images will loss the contrast and color artifacts from the original image. Edge detection is another challenging issue on such kinds of degraded images. There are several research works are under progress to reduce the haze exists in the image. Although haze removal techniques will reduce the haze present in the image, the results of those techniques were dropped the natural look of the original image as penalty. We proposed an effective way of finding the edges from the hazy images. Firstly, a dark channel prior method is used to eliminate the unwanted haze from the original image. The statistics shows that this method effectively works for the images taken in an outdoor hazy environment. The key observation of this method is that at least one color channel is having a minimum intensity value in a local patch. The results shows that results of this method have a good results compared to other contrast improvement techniques. Secondly we have applied the Sobel edge detection operator to find the edges of the resultant image.

Article Details

How to Cite
, Y. S. I. M. S. (2018). An Efficient Edge Detection Technique for Hazy Images using DCP. International Journal on Future Revolution in Computer Science &Amp; Communication Engineering, 4(5), 12–18. Retrieved from http://ijfrcsce.org/index.php/ijfrcsce/article/view/1637
Section
Articles