2001 Volume 21 Issue 6 Pages 407-415
The authors previously developed the tele-rehabilitation support system. Occupational therapists (OT) and Physical therapists (PT) reported that the system was useful for supporting general rehabilitation training. However, they also pointed out its insufficiency in recovery evaluation, because of the incapability of motion angle detection. Therefore we have developed a new method for a motion angle measuring system from training tele-video images with a one camera imaging system, in order to evaluate the rehabilitation outcome.
Our method consists of six steps of image processing, Discriminant Analysis Binarization, Hilditch Thinning, Hough Transform, Cluster Analysis and Angle Detection. This sequential processing automatically detects both centerlines of forearm (basic axis) and of hand (movement axis), which made a centerline in the hand. Also, this basic axis and the movement axis were classified by cluster analysis.
The result shows the new method has enough capabilities to detect the range of motion in training-effect with precision. We concluded that this method will be very useful for tele-rehabilitation, when angle detection errors in this method are finely corrected.
