Edge based Crime Assistance System with Cloud Computing

Abstract

Nowadays, the level of crime is increasing more and more in every corner of the world. With the rising rate of criminal activities, the identification process of the person who committed crime takes a lot of time. Most of the crime reporting systems currently are reported personally. In order to create a fast and reliable crime reporting system, the face identification in crime assistance system between edge and cloud computing is proposed. The processing power of ubiquitous devices (e.g. smartphones, sensors, and actuators) are very powerful today. They allow capturing the image of the criminal anywhere at any time. They also have the capability of detecting faces from captured image. Therefore, the face detection process using Haar face detection and face identifier generation process using Local Binary Pattern (LBP) from the captured image can take place on the edge devices. Face identifier matching is proceeded on the cloud. The paper proposed the edge-based face identification system and represents the analysis result of face detection time, matching time and total processing time of the proposed system with the increasing number of test image data set. The results show the effective usage of edge devices and improve the efficiency of face identification.