Outdoor Long Range Ubiquitous Projectiles Tracking System Using P-MPLR and Computer Vision

초록

A growing body of research has been dedicated to investigating systems for quickly detecting and pinpointing the location of projectiles aimed at a target. Various studies and products have demonstrated the feasibility of predicting the point of origin for a shot. However, existing tools come with limitations and inconveniences. This study aims to introduce a model that achieves highly accurate and low-latency tracking of shot origins during long-range shooting, free from constraints, by incorporating LoRa, computer vision, and sound detection technologies. The proposed system commences with an initial photograph captured by a camera positioned near the target. Whenever the adjacent acoustic sensor detects a fired projectile, the camera captures a new image, triggering the Raspberry Pi located beside the target to calculate the coordinates of the shot by comparing it with the previous image. When compared to the actual bullet mark, the predicted bullet mark's error remains under 1 mm. This paper's proposal showcases accuracy suitable for integration into military intelligence training systems, offering a cost-effective solution. © 2023 IEEE.

키워드