mohammadmozafari / face-clustering

This is the implementation of my BSc project. A software that can group pictures based on the identity of the people in them.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Project Description

The goal of this project is to build an intelligent system that performs clustering on face images. Face clustering is used in mobile phones to categorize gallery images. Such a system can also be used to review CCTV videos. This system is implemented in the form of a desktop software. The software has a graphical interface suitable for communication with user and consists of several major parts. In the first part, the user identifies the folder where there are a number of image. The program finds all pictures in this folder and presents them to the user, then it processes the images and finds every face image in them. There are plenty of algorithms that we can use for face detection e.g. MTCNN which is used in this project. After detecting the face images, another algorithm is used to find a representation for each face. This representation is later used in a clustering algorithm and images are categorized according to the identity of the people in them. Finally images of different people are saved in separate folders and the user can view them.

home home home

About

This is the implementation of my BSc project. A software that can group pictures based on the identity of the people in them.


Languages

Language:Python 98.2%Language:Shell 1.8%