Chapter 3: Segmentation

Title: Histogram based image segmentations using connectivity analysis

Author(s): Indrani Chakravarty, Pabitra Mitra

Address: Department of Computer Sciences and Engineering, Indian Institute of Technology, Kanpur, India | Department of Computer Science, Indian Institute of Technology, Kharagpur, India

Reference: Atlantic Europe Conference on Remote Imaging and Spectroscopy pp. 83 - 90

Abstract/Summary: Colour images have three components – red, green and blue. All the three planes can be combined together to form an intensity image, which can be further subdivided into different regions based on the intensity values. However, two distinct regions without any spatial connectivity, but having the same intensity value, can be considered as part of one contiguous region. Application of a Depth First or Breadth First Search will help in distinguishing between two such regions. But, a direct application of the search, without any pre-processing (merging), on the input image runs the risk of labelling two parts of the same region to be labelled differently. In this paper, an algorithm is proposed to segment colour or intensity images using an initial merging, followed by a splitting technique, which does a connectivity analysis on the merged image using information from the histogram. Application of this algorithm lies very much in biomedical and remote image processing.

Order a copy of this article