In authenticating and verifying important documents, one of them is in the form of identifying the authenticity of a signature. In addition, the signature is also a form of ratification and a sign of approval in important documents is mandatory. Along with current technological developments, the signing process can be carried out in digital media such as cellphones and other media. The ability of the system to identify a person's signature becomes important because of the many forgeries that occur. This study aims to implement the Canny edge detection method to identify a person's signature. The number of signature images used is 10 signatures. The results of this study indicate that the Canny edge detection method has a similarity percentage of 70% to 100%, and the similarity values below 70% and above 100% are grouped into signature images that are not original.