This project implements a real-time embedded system for detecting and analyzing shots fired at a shooting range, combining Digital Signal Processing (DSP), ultrasonic radar, and a neural network classifier optimized with genetic algorithms (NNGA).
Built on a Raspberry Pi Pico, the system samples distance and motion data using the HC-SR04 ultrasonic sensor and MPU6500 accelerometer/gyroscope, applies IIR-based signal smoothing, and visualizes results in real time. The classified output helps assess shot placement accuracy and improve user performance. Despite hardware limitations, the system achieves sub-100 ms latency and stable classification on modest microcontroller hardware.
🛠 Technologies: Embedded C/Python, real-time DSP, UART, Matplotlib, NumPy, NNGA
🎯 Goals: Real-time shot tracking, performance feedback, low-cost and portable radar
📡 Features: Signal filtering, polar plotting, feature engineering, softmax classification
This work was presented at the 2025 Sesiunea de Comunicări Științifice Studențești UPB and earned an honorable mention for electronics and real-world applicability.