Abstract:
In recent times, the demand for clean urban environments has grown significantly due to considerations for the environment and public health. However, littering remains a common problem, especially in crowded places, leading to pollution and affecting the quality of life. As a solution, we present LitterCam AI, a system employing state-of-the-art computer vision techniques to detect vehicle based littering offences and identify the vehicle license plate involved. Using advanced algorithms, the littercam AI searches and monitors vehicles with the nearest car littered or involved in littering within any urban setting. The system was developed from neural networks trained with the very fast and accurate YOLOv8 architecture, which recognizes vehicles and rubbish on dynamic images. Once littering is reported, the system simply takes a picture of the vehicle, especially targeting its license plate. Then, the image is processed by one of the most common tools for optical character recognition named EasyOCR to extract the license number for further actions. LitterCam AI is a stride towards environmental conservation through artificial intelligence. It intends not only to dissuade the littering habit but also create awareness for clean public spaces. Thus, by adopting technologies and a fresh approach to the problem could bring us a future that is cleaner and more sustainable.