A framework for edge detection based on relief functions.

Abstract: In this work we introduce a novel edge detection framework based on the measurement of grey level changes using a new class of functions called relief functions. We first analyze such functions and propose a construction method based on binary functions. We also study several properties of the binary functions, how they affect the relief functions and their application in our edge detection framework. Moreover, we compare the results obtained by the proposed framework with those by the Canny method on both general-purpose and problem-specific images. The experimental results show that the methods embodying the proposed framework using relief functions are competitive with the Canny method, while providing higher flexibility. [Copyright &y& Elsevier]