Abstract: Edge detection finds extensive applications in fields such as autonomous driving, pattern recognition, and biomedical imaging. Optical edge detection is particularly favored by researchers due to its inherent advantages of high speed and parallel processing capabilities. Metasurfaces, capable of precisely manipulating light fields within an extremely compact subwavelength space, have been widely applied in optical edge detection systems. This review centers around the two primary design methodologies for optical edge detection based on metasurfaces. It introduces their fundamental principles and characteristics, summarizes relevant research progress both domestically and internationally, and proposes promising future development directions.