A measure of quality for evaluating methods of segmentation and edge detection

A new measure of quality is proposed for evaluating the performance of avai lable methods of image segmentation and edge detection. The technique is in tended for the evaluation of low error results and features an objective as sessment of discrepancy with respect to the theoretical edge, in tandem wit h subjective visual evaluation using both the neighbourhood and error-inter action criteria. The proposed mathematical model is extremely simple, even from the perspective of computational execution. A training of the measure has been put in practice, which uses visual evaluation of a set of error pa tterns by a team of observers. Encouraging results were obtained for a sele ction of test images, especially in relation to other recently proposed and /or currently employed quality measures. (C) 2001 Pattern Recognition Socie ty. Published by Elsevier Science Ltd. All rights reserved.