Fractional Order Sobel Edge Detector

Abstract

Edge detection is widely used in different applications in image processing and analysis. It is a relatively old concept that relies on numerical derivatives to detect image edges, mainly of the first (gradient) or second (Laplacian) order. With the emergence of fractional calculus, edge detection is revisited with non-integer order derivatives. This paper proposes the design of a fractional order Sobel edge detector. Sobel gradient operators are considered for the first order derivative, and fractional calculus is exploited for non-integer orders beyond unity. The proposed method is implemented and assessed in the context of road obstacle detection, and contrasted with the conventional Sobel edge detector.