Design and Implementation of a Control Device for Intelligent Wheelchair by Combining Image Processing and Acceleration Sensor

In this paper, a gadget for controlling an electric wheelchair (EWC) is designed.The device is designed based on the combination of acceleration sensor data fromhead rotation, and image processing data from user’s face recognition, forcommanding to the EWC. This gadget designed as a wearable device and isdeveloped low cost, safe, and flexible for the patients with spinal cord injuries aswell as the elderly with limited hand use. The mechanical design, sensor tuning,and 3Dprinted prototype of the gadget are presented. Finally the result ofexperimental test is discussed. Acceleration sensor module and camera andRaspberry Pi board are the core of the gadget. For performance evaluation,several experiments have been performed. The commands sent by the user aredivided into four control commands (right, stop, move, left). Forward movementcommand, is performed by showing a happy expression on the face. Theexperiments show that the head angle controller and the image processor reactto the stop in the shortest time, which indicates that it has a high level of safety.