Development of Vision Guided Pick-n-Place Robotic Arm

This book presents a complete vision guided robotic arm system for picking and placing of objects. For object recognition and localization, the approach exploits Scale Invariant Feature Transform keypoint extraction to segment the correspondences between the object model and the image onto different potential object instances with real-time performance. Form the obtained correspondences, the best picking point is estimated. The coordinates of this point is then transferred to robotic arm coordinate system, which allows the arm to pick and place the object to the desired locations. The use of SIFT based clustering allows the system to be used for applications under extreme conditions of occlusion, where standard appearance-based approaches are likely to be ineffective. This system overcome most of the challenges occurred in actual workspace and in real time applications by presenting sufficient speed of operation. The system can handle complex ambient illumination conditions, challenging specular backgrounds, condition of occlusion, diffuse, and transparent as well as specular objects efficiently. The system can work with a single view of object, reducing the time and complexity.