Design, Control, And Development Of An Intelligent Of Waste Sorting System With A Robotic Arm

Abstract

The paper has presented the control and development of an intelligent garbage sorting system with a robot arm. This system consists of a machine vision block, a 6DOF robot manipulator, and a control unit for sorting garbage based on analytical images. YOLOv4 software will be identified the object by the neural network. This method is used for detecting and image recognition of different sizes and types of waste, such as paper-based garbage, metal garbage, and plastic garbage. The results of offline testing on a database of more than 600 untrained images show that the trained model has an average accuracy of about 98.43% for classifying different types of garbage.

Keywords

• waste sorting
• robotic arm
• yolo
• machine vision
• 6dof