Face Recognition and Weapon Detection from Very Low Resolution Image

Abstract

In this paper, we are implementing a convolutional neural network for face recognition. We preprocess the pictures by utilizing channels. Haar cascade is utilized to distinguish face and weapon. Convolutional neural system is utilized upon the distinguished pictures to prepare the system for perceiving the pictures. CNN utilizes less pre processing when contrasted with other image classification algorithm. CNN utilizes neurons which naturally learns based on weights and bias. For any new faces discovered we will catch the arrangement of pictures of the new face and prepare them under a name. With the goal that new faces likewise can be trained and perceived utilizing CNN.