CRIME PREDICTION SYSTEM USING DEEP LEARNING AND CONVOLUTIONAL NEURAL NETWORK: A SYSTEMATIC REVIEW AND FUTURE ENHANCEMENT

Abstract

Over the past decades, many machine learning approaches have been proposed to identify crime activities from inertial sensor data for specific applications Most methods, however, are designed for offline processing rather than processing on the sensor node. In this project, a crime prediction technique based on a deep learning methodology is designed to enable accurate and by using CNN-real-time classification using video processing by python. To obtain invariance against changes in human movement, motion, feature extraction and we design a feature generation process that is applied to the spectral domain of the inertial data. Specifically, the proposed method uses sums of temporal convolutions of the transformed input. Accuracy of the proposed approach is evaluated against the current state-of-the-art methods using both laboratory and real world activity datasets. A systematic analysis of the feature generation parameters and a comparison of activity recognition computation times on any device are also presented. This system mainly focuses on the prediction of the crime by CCTV footage and provide the type of crime, also generates the report based on crime location and time.