PLEESE - Plastic Reuse Reminder System using Computer Vision Deep Learning

Plastic pollution is a well-known problem but very challenging to solve. A new paradigm, the circular plastics economy, has been introduced recently to address this problem. Reducing, reusing, recycling, and composting are four essential aspects of the circular plastics economy. However, most industries and societies focus solely on reducing, recycling, and composting. Thus, here we developed PLEESE, a plastic reuse reminder system that aims to raise people's awareness to reuse plastic into something of economic value. PLEESE was developed using computer vision and deep learning. The system uses a Jetson Nano single-board computer, digital camera, and a display monitor mounted above the trash can to detect plastic bottles automatically. Detection of plastic bottles was carried out using SSD-MobileNet-V1 as the deep learning model. Both hardware and software used in this work are described in detail with the hope that this system can be implemented widely in strategic waste areas soon.