A rehabilitation training system with double-CCD camera and automatic spatial positioning technique

This study aimed to develop a computer game for machine vision integrated rehabilitation training system. The main function of the system is to allow users to conduct hand grasp-and-place movement through machine vision integration. Images are captured by a double-CCD camera, and then positioned on a large screen. After defining the right, left, upper, and lower boundaries of the captured images, an automatic spatial positioning technique is employed to obtain their correlation functions, and lookup tables are defined for cameras. This system can provide rehabilitation courses and games that allow users to exercise grasp-and-place movements, in order to improve their upper limb movement control, trigger trunk control, and balance training.

Research Highlights
► A special human–machine interface is developed by integrating two CCD cameras and a large computer screen, and linked the rehabilitation equipment to the interactive game.
► An automatic spatial positioning technique is used to obtain the correlation functions and lookup tables of two cameras.
► The data were computed and recorded into a training performance database. The record consisted of maximal movement velocity (MV), movement response time (MT), and training game score.