In this work we present an automated parking lot system that is managed and monitored by a video camera. We used a web camera for monitoring and the PC with MATLAB environment for managing the parking lot model. Captured frame is first converted to the grayscale image and then to the binary image with Sobel edge detection method. Image is dilated and filled in order to get the cars in white pixels and the parking place as black pixels. To each parking place is dedicated a certain region in the image. Mean value of each region was used for indicator if the parking place is occupied or free. By using a camera as a sensor, relatively large parking area is covered with only one sensor. Proposed system is relatively cheap. It does not require major additional installations on the existing parking lot. Real-time experiment was made on a parking lot model. Experimentally observed execution time implies to a possible application on a real parking lot system. |