International Research Journal of Education and Technology

In this project, we have developed a system for detecting criminal faces, for this, we have used deep learning algorithms. Deep learning is currently the most wellknown technique, and it is utilized in a variety of applications. One such application is crime detection and prevention. The criminal's face is recognized by this system, which then obtains the criminal's database information. In this application detect image and video for criminal identification. Once select the image to identify it results whether the person is criminal or non-criminal. It will simplify the investigation process. We use pre trained convolution neural network architecture called VGG-16. It can classify the images among 1000 categories. We collected our datasets from NIST special datasets 18. This project will identify whether the person is criminal or non-criminal. It can able to run in a real time. this project can able to process video and detect the identity of the person through video only if the person is criminal. Otherwise, recognize as non-criminal.

Key Words: Deep learning, Criminal detection, Real time application, Image recognition, Video detection