Subpixel edge detection of color images by principal axis analysis and moment-preserving principle

Abstract

This paper discusses a new approach for detecting edges with sub-pixel accuracy in color images using the principal axis analysis and the moment-preserving principle. First, the principal axis of the color vectors in a three-dimensional space is obtained analytically for each image block. Based on the projection scores, the direction of the line edge is obtained by analyzing the three-dimensional principal axis of a color image block. Next, by preserving some spatial color moments of the pixels of each square block, analytic formulas are derived to extract the edge feature. Experiments on both synthetic and real images show that the accuracy of the proposed method is comparable to other color edge detectors. The proposed algorithm can be performed very fast for real-time applications with no need for special hardware.