Because the gray level and contrast of infrared images are lower than those of visible images, it is not effective for an edge detection method based on gray level to detect infrared images. For this reason, an infrared image edge detection method based on phase consistency is proposed according to the invariant principle of phase consistency for gray scale and contrast. First, the phase consistency value of an infrared image is calculated. Then, the regions containing edges are segmented by K-mean's method. Finally, the edge curves are generated by a morphological method. The experimental results show that the proposed method is more effective than the Canny algorithm in infrared image edge detection.