Automatic Color Edge Detection with Similarity Transformation

Abstract

Edge detection is an important step in image processing. As edge is intensity variation with spatialcoordinates, the similarities between neighboring pixels could be used for edge detection. It hasbeen observed that the effective results could be attained by thresholding the homogeneity imagesgenerated by means of the similarity transformation. Nevertheless, the user-defined normalizationcoefficient in similarity transform stage seriously effects edge detection performance and it needsto be automatically selected for every particular image. In this study, a new approach in whichthe normalization coefficient is automatically determined has been presented. The automatingprocess of the similarity transform has been performed according to the gray level values of theneighboring pixels. The gray level differences of the central pixel and other neighboring pixelshave been used to determine the similarity coefficient. Subsequently, the binarization process ofthe homogeneity images obtained with proposed algorithm have been completed with differentthresholding techniques. Additionally, the F-score of the proposed edge detection has beenobtained with 200 images in the BSDS training dataset. The achieved F-score values have showedthat the performance of automatic approach is quite high.