Design and implementation of an intelligent waste classification system using computer vision

Development of an intelligent waste classification system for automated detection and management of recyclable materials. The system uses a Raspberry Pi 5 running a YOLO-based computer vision model to identify waste in real time and determine whether an item is suitable for recycling. The project integrated computer vision with electronic control and power systems for the mechanical components of the prototype. Power supply, motor power, and control circuits were designed and implemented to support the operation of the system. A graphical user interface running on the Raspberry Pi was developed to display the amount of recycled waste and its corresponding credits for the user, as well as information related to the waste classification process. The project was developed in collaboration with the PUCP Robotics Group and the University Social Responsibility group of the Pontifical Catholic University of Peru, from July to December 2025.

System Implementation

Power and Motor Control Electronics

Electronic control and power circuitry developed to supply the Raspberry Pi and camera, as well as to drive the motors responsible for the mechanical operation of the waste classification system.

User Interface

Graphical user interface developed for the Raspberry Pi display, allowing users to interact with the system, access information about the recycling process, and visualize the amount of recycled waste and the corresponding credits assigned to the user.

Final Prototype

Final implementation of the intelligent waste classification system, integrating the Raspberry Pi 5, camera, mechanical structure, control electronics, and internal wiring into the complete prototype.