Finding Connected Components in a Gray Scale Image

Abstract

Abstract—

Finding connected components are well defined for binary images. The concept of connected components can be extended for gray level image. But the problem is the criteria based on which a connected component would be defined. A gray level image is an image having 256 different pixel intensity levels. If we consider connected regions having only a particular pixel values, the number of connected components would not be meaningful and the purpose of finding connected components would be lost. So, we define a connected component in a gray scale image based on range of pixel mapping and new method to find connected components in a gray scale image is proposed. Three different types of pixel range mapping are introduced, using connected components in a gray level image can be successfully found. Connected components in a gray level image are the segments of image having the same range of pixel values. Different regions or segments of image can be found easily.

Keywords—Connected component, gray scale labelling, pixel range mapping, linear mapping, logarithmic mapping, square root mapping.

(Article history: Received 1 November 2016 and accepted 30 December 2016)