Development of Student’s Uniform Compliance Detection System Using Real Time Image Recognition at Politeknik Statistika STIS
Keywords:computer vision, object detection, YOLO, system development
Regulations in the Politeknik Statistika STIS (hereinafter called Polstat STIS) aims to produce graduates who are qualified, with integrity and trusted. In enforcing regulations in Polstat STIS, there are a student squad of regulations enforcement, which is called Satuan Penegak Disiplin or SPD in Indonesian, which aims to maintain the order, discipline and student ethics during their activities on and off campus. In upholding the regulations, SPD carries out surprise inspection and during the weekly morning assembly to check completeness and tidiness of student’s uniform as well as his/her look. However, previous research related to the student’s commitment to the campus regulations shows that half of the students have low commitment. This is partly due to the lack of supervision of students. Therefore, it is necessary to monitor the discipline and neatness of students on an ongoing basis. In order to conduct the monitoring and inspection on a more regular basis, the method of image recognition can be used to assist in overseeing student discipline and neatness. In this study, we developed a system which can detect in a real-time manner the completeness of attributes the student wears. The system we developed uses object detection to detect the completeness of student attributes. The system shows and records student(s) whose attributes are incomplete. The system expected to improve the discipline and neatness of the students.