Deteksi Ketersediaan Lahan Parkir Mobil Menggunakan Yolo V4 Berbasis Website

Abstract

The car park availability detection system addresses the challenges of monitoring and managing parking lots that are often overfilled. This research focuses on designing a detection system using the YOLOv4 algorithm, integrated with a web-based platform. The system utilizes deep learning-based object detection to identify vehicles in parking areas. Users can upload images or videos of the parking lot through the website, which are processed by the YOLOv4 model to detect parked cars and determine parking space availability. The interface highlights empty spaces in green and occupied ones in red, making it easier for users to identify available spots. The model was developed using 130 training data samples and tested with 33 data samples. It achieved a remarkable Mean Average Precision (mAP) of 100%, demonstrating its high detection accuracy. This innovative solution is expected to enhance parking management efficiency, enabling users to access real-time information about parking availability conveniently and effectively.