Image Edge Detection Based on Contourlet Transform Combined with the Model of Anisotropic Receptive Fields

The existing algorithms of image edge detection based on space domain can effectively detect the edge of the image in limited direction. In order to solve this problem, according to the direction information and gradient direction information obtained from each directional sub band of the Contourlet transform subband, this paper proposes a new algorithm of Based on Contourlet transform combined with anisotropic receptive field model of image edge detection. This algorithm firstly carries Contourlet transform on the original image, and detects the edge on each scale. It compensates and corrects the image edge with high and low frequency respectively. Finally, the edge image is obtained by choosing the coefficient of absolute value maximum. The experimental results show that this algorithm is a new way with continuous edge and accurate positioning, which reduces the false edge.