DEVELOPMENT OF AN ARTIFICIAL VISION PLATFORM FOR VIRTUAL SHOOTING TRAINING THROUGH OPTICAL DETECTION OF LIGHT BEAMS

Abstract

This article presents the development of a computer vision platform for virtual shooting training, using optical light beam detection as a mechanism for evaluating accuracy and performance. The system was designed to simulate school shooting instruction scenarios in controlled environments, significantly reducing operating costs and increasing the frequency of practice. Through the use of OpenCV and AForge.NET libraries, it was possible to detect with high precision the impact of laser beams projected onto virtual silhouettes, integrating the information into a database for automatic performance evaluation. This platform aims to strengthen the instruction of first-semester students, significantly reducing the operating costs associated with physical practice and increasing the frequency and quality of training sessions. The results demonstrated 90% accuracy and a 75% reduction in false positives compared to the color-based detection method. This development demonstrates the potential of computer vision applied to military training and its viability as an interactive and scalable instructional tool.