Photo Automatic Classification System Based on Face Recognition

摘要

It is very important to manage a large number of digital pictures for photography enthusiasts and training institutions. There are time-consuming, labor-intensive, and low-efficiency for classification using manual classification methods. They are used to categorize photos based on their shooting time or location when managing photos. It is difficult to find photos that can be of interest to the customer. Photo automatic classification system based on face recognition is proposed, which is convenient for users to find their own photos. Given a single photo of a person, the system first conducts self-deep learning to obtain the feature value of the face. And then the system can find all the photos in the given folder and copy the found photos including the sample face to the corresponding person name folder after reading the sample photo. At the same time, photos that cannot be recognized are also copied to the undetected face folder, so that the user can do a small amount of manual classification.