PARK VISION: INTELLIGENT PARKING MANAGEMENT SYSTEM
Over the past several years, there has been an increase in the number of automobiles in use. Larger parking lots are therefore clearly needed. The current standard techniques for identifying if a slot is occupied in smart car parking lots are no longer appropriate as they require a large number of costly sensors and the region that required observation growing. The purpose of this study is to determine, update, and display the current number of open parking spaces in the parking lot using an efficient, quick, and accurate method. A camera's worth of video was employed as the input medium, and the object identification technique for image processing was Yolov5. By comparing the independently discovered locations of parking lots and parked cars, available parking spots were assessed. The dataset utilized to train and assess the model was the PKLot database. The photographs included in the dataset that corresponded to various weather conditions were used to assess the performance of the suggested model. The model's performance was 90.03% on average. Bright days showed the best performance, and rainy days produced the lowest.
Convolutional Neural Network, Computation of Images, Shortest Path Method, Smart Parking System, and Parking Spot Recognition.