Traditional live-fire training faces challenges such as high costs and risks, while existing simulation training systems suffer from high costs and complex hardware. This study develops a lightweight simulation system for shooting training, achieving low-cost and high-precision training through hardware integration and algorithm optimization. At the hardware level, a lightweight architecture integrates simulated weapon, image acquisition devices, and a processing core. At the software level, a spot recognition framework based on the Two-Pass algorithm is adopted, incorporating adaptive threshold adjustment, area filtering, and circularity detection as post-processing steps to enhance robustness. Additionally, the system uses the UDP protocol for real-time data transmission and perspective transformation for image calibration. Experimental results show that the system achieves a spot recognition rate exceeding 99% in indoor environments, with the Two-Pass algorithm-based positioning method demonstrating 0.12-pixel accuracy. In multiple simulated shooting tests, this system demonstrates excellent stability and reliability, providing an innovative and practical technical solution for simulation shooting fields.