Project Overview
This smart parking management solution leverages Computer Vision edge processing to monitor vacancy/occupancy dynamics. By processing live camera streams over RTSP feeds, custom models segment parking boundaries, detect vehicle bounds, and update slot status indexes instantly.
Accuracy & Features
The detection model reaches a verified 96% accuracy rate across 8 active monitored sectors. Features include a dynamic administrative dashboard displaying real-time statistics, vacancy heatmaps, automated parking slip logging, and database triggers that push status indices to driver navigation interfaces.
Accuracy
96% verified detection accuracy across 8 lots.
Tech Stack
Python, OpenCV, YOLO model segmentation, Flask APIs, SQL Databases, React Dashboards
Role
Lead CV & IoT Engineer