Sistem Identifikasi Pergerakan Mencurigakan Peserta Ujian CBT Politeknik Caltex Riau Menggunakan IP Camera

Abstract

One surveillance system that is often used is CCTV. This CCTV will provide results in the form of video recordings which can then be accessed by wireless communication. In its use, CCTV must still be protected by human resources to give a warning. Then the use of CCTV becomes less efficient when used to oversee a room where the room rarely has movement. Because CCTV cannot detect or identify suspicious movements automatically. So we need a system that can identify movements (suspicious movements) automatically. In this case, the movement to be identified is the movement towards the Caltex Riau CBT Polytechnic exam participants. Monitoring cameras (monitors) which are also often done to oversee an area or object, one of which is the IP Camera. Based on these problems, this system will be developed using IP as an input image. Then the image will be carried out a feature extraction process. Then enter the classification process using the k-Nearest Neighbor algorithm to determine the label pattern of the examinees' movements. Based on tests that have been done using 502 data, the accuracy rate obtained is in the range of 98% - 98.4% with an average accuracy of 98.2% for the value of neighborliness = 3. With a system of 98.2% system accuracy can be used to identify suspicious movements of CBT Polytechnic Caltex Riau exam participants using IP cameras, where the movement conditions are not suspicious or suspicious movements.