Synthetic Aperture Radar（SAR）image edge detection is a hot issue in the field of image processing. It is difficult to eliminate the noise and internal texture of SAR image by traditional methods. By making use of the properties of the receptive field and the surround suppression mechanism to texture and noise in the homogeneous region in cells with nonclassical receptive field in Primary Visual Cortex (V1) in human visual system, this paper proposes a new method of SAR image edge detection by modeling the information processing mechanism of V1 Through the operations of odd and even decomposition, spatial summation and normalization by Gabor filter, the paper simulates the properties of the classical receptive field. A series of computational steps motivated by surround suppression mechanism are added to suppress the texture and noise in homogeneous region and preserve the contour wanted. The simulation results show that the proposed method can suppress most texture in the homogeneous region of the four kinds of SAR images used in the article，can weaken speckle noise to some degree, and edge detection effect is better than that of the canny detector. The method can be applied to the edge detection of SAR images.