Parking Occupancy Detection using Computer Vision Techniques

The surge in vehicle numbers on roads contributes significantly to traffic congestion and management challenges, particularly evident in developing nations like India where the influx of cars exceeds road and parking capacity. Addressing these issues necessitates the implementation of sophisticated parking management systems. This project focuses on two key objectives: detecting vehicle occupancy within marked parking slots and analyzing parking data. Using the parking lot near IIT Kanpur main gate as a reference, video data was collected for 14 consecutive days, enabling the evaluation of vehicle occupancy and parking patterns. Object detection algorithms such as Mask-RCNN and YOLO-v5 were employed to identify occupied parking spaces within the lot. Various methods, including HAAR cascade-based classifiers, DNN-based systems utilizing ResNet classifiers, and RCNN with IoU, were tested for detecting vehicles within allotted slots. The data collected was stored in CSV format for analysis. This project aims to provide insights into detecting parking space availability and analyzing parking data to optimize time and fuel efficiency. In the Mask-RCNN approach, pre-occupied spaces are denoted by red boxes, while green boxes represent available parking spots. Similarly, YOLOv5 was utilized to count cars in video frames and identify available parking spaces. The YOLO Annotation Toolbox facilitated the extraction of parking space coordinates from recorded video frames, which were then visualized in QGIS for further analysis.