Trespassing Detection System on a Smart Video Surveillance System for Level Crossing

The Ministry of Transportation of the Republic of Indonesia recorded 1,142 cases of accidents at level crossings from 2019 to 2022. Currently, the surveillance system at level crossings still conventional. CCTV cameras, which are only capable of monitoring, have not been able to effectively solve problems at level crossings. Therefore, there is a need for technology that can enhance the capabilities of CCTV cameras in tackling these problems, one of which is by applying computer vision techniques. In the past decade, research has been conducted on video surveillance systems with a computer vision approach. Most of this research aims to replace human functions or conventional systems with video processing algorithms. In this paper, a trespassing detection system was developed for smart video surveillance to enhance monitoring at level crossings. The implemented system can understand the situation that occurs at level crossings and can accurately identify objects of tresspasing at level crossings, the accuracy rate reaches an of 86%.