This thesis represents a smart parking management system designed to enhance parking space utilization, which was designed for a private parking lot company in Bosnia and Herzegovina. The system leverages image processing integrated with CCTV cameras to detect and monitor parking space availability in real time. The primary objective of the system includes accurately detecting free parking spots. Through image processing algorithms, the system analyzes CCTV footage to identify non-zero pixels, indicating the presence of vehicles in parking spaces. Later, the system displays bird eye view of parking spaces and through unique identifiers displays car icons on taken parking spots, which is displayed at the entry terminal of the parking lot. In this manner, the drivers get the information whether there is a free parking spot in the parking lot or not and its location.
In the first part of the thesis, we explain theoretical parts, which are needed to understand the performed project described. Afterwards some of the image segmentation techniques are explained and compared, which are used to detect parked vehicles. Finally, a complete project is presented, including a demonstration of its performance.
Initially, the project’s goal was to detect the availability of parking spaces but can easily be upgraded, which is explained at the very end of this thesis. This project was made through a combination of Python and OpenCV library and image processing techniques, which enabled us to make a simple yet effective solution to this problem.
|
