Automatic Chromatic Aberration Detection using Local Properties

Abstract

Chromatic aberration is introduced by numerous rays of light travelling through lens out of convergence
point onto image sensor in digital camera. It causes the colors to be blurred on the image. It decreases the
quality of images and also has annoying effects as well. It is hard to automatically detect the color aberration.
Therefore, the aberration area is hardly identified because it is hard to identify the incident point of each color on
the lens. To the best of our knowledge, automatic detection method has not been proposed. This paper therefore
proposes a method to automatically detect chromatic aberration in images. By incorporating edges with local
properties, such as contrast and threshold, the aberration area can be detected with high proximity in each area
of images. The experimental results show that the proposed method can positively discriminate natural color area
and chromatic aberration area efficiently. Moreover, the accuracy of the proposed method is 90.11 percent by the
average.