Speech Driven Robotic Arm for Sorting Objects Based on Colors and Shapes

Automation has played a significant role in assisting humans by relieving them of tedious, routine and mechanical tasks in their day to day life. Robotic arm has been used in industry to replace human arm to perform picking, moving, placing and sorting the manufactured objects. This paper is an attempt in this direction to provide a novel speech enabled robotic arm which can sort objects based on colors and shapes. It consists of a 4-DOF robotic arm whose kinematics are controlled by an Arduino microcontroller. The speech recognition feature is implemented in Matlab running on a Windows laptop, which uses speech processing algorithms from the DSP toolbox. The objects are identified by images captured by the camera and sent to Matlab for color and shape recognition. Color recognition is implemented based on the RGB components extraction, whereas the shape recognition is achieved by measuring the circularity of the object shape. The novelty of the implemented solution lies in providing the desired functionality of object sorting which is economical, small-sized, accurate and uses popular programming tools. This solution can be potentially used for both industrial and educational purposes.