Akademik Veri Yönetim Sistemi
Sevri M. (Yürütücü), Güneyli M., Özkan M.
Yükseköğretim Kurumları Destekli Proje, 2025 - 2025
Today, the increasing need for raw materials and environmental pollution have become one of the most important problems for humanity. In this context, the RecyTech project aims to optimise waste sorting processes using object detection and deep learning based image processing techniques. In the project, lightweight and fast deep learning architectures such as Single Shot Detector (SSD) and MobileNet will be used to classify waste with high accuracy in real time. Waste images will be captured by means of high-resolution cameras and embedded systems to be integrated into recycling bins, processed by a PyTorch-based model, and object types will be identified and directed to the appropriate compartments. Within the scope of the project, automatic sorting of common recyclable materials such as plastic, glass, metal and paper will be provided. In addition, thanks to the mechanical sorting mechanism supported by smart servo motors and opening doors, it is aimed to realise the recycling process fully automatically without the need for manual intervention. In the second phase of the project, a mobile application integrated with recycling bins will be developed to increase user interaction; this application will be supported by gamification techniques and reward systems to encourage users' recycling habits. At the end of the project, it is aimed to significantly improve recycling processes in terms of accuracy and efficiency compared to traditional methods, increase environmental awareness and sustainably increase recycling rates.