Object recognition with camera images on moving conveyor and pick and place application the objects

Nowadays, image processing applications are increased and as paralel, the application areas of image processing implementations are developed in wide areas. Embedded imaging tools like cameras and the program algorithms had been begun to use like a sensor and astandalone machine. The goal in this study, is to evaluate the success of the study concerning the determination of the material on conveyor from the production machine to packaging. It is expected that this determination should have some features like determination the faulty material and recognition the correct product.In this thesis, it is aimed to recognize, classify without regarding its location, orientation in an image and to pick and place the target parts with the robot arm. For this aim, an experimental system is designed and implemented. This system includes an active camera system, the image processing algorithm which is used in feature extraction, an Artificial Neural Network (ANN) which is used for recognition and classification, and a robot arm and its control unit for pick and place application.In the developed system, a speed control circuit for conveyor, a control unit with a microcontroller for communication with computer and a robot arm have been applied. The serial robot manipulator arm has three axis and uses RC type servo motors. The robot arm body is laser-cut aluminium material. The machine vision has been provided using a webcam camera which is placed in the conveyor box.It is hoped that application of this thesis can be used in development of similiar systems and these type systems can be used in the production lines.Keywords: Object recognition, robot arm, image processing, neural network