Robotic arm for pick & place operation using matlab based on offline surface clustering algorithm

The industry is moving from current state of automation to Robotization, to increase productivity and to deliver uniform quality. The main objective of this paper is to perform an object detection and its pose estimation for robotic pick and place operation using offline surface clustering algorithm. The pick and place robot is a microcontroller based mechatronic system that detects the object, picks that object from source location and place it at desired location. RGBD Sensor provides depth information beyond visual data. In the present situation the detection part is mainly based on the recent template-based linemod approach for object detection. Offline surface clustering algorithm was introduced to overcome the difficulties that occur in present situation and it improves the correct detection rate compared to linemod approach, hence suitable for robotic applications. Object detection by image processing algorithms is done using MATLAB software. Hardware implementation of robotic pick & place task for detected object is done using PIC microcontroller.