Mohammad Kavian
Ali Nadian-Ghomsheh

Machine vision techniques are gaining more attention for monitoring the performance of physically disabled patients. Such monitoring systems lead to diagnosis and treatment plans to improve the life quality of patients. In this paper, vision based method for wrist and fingers range of motion measurement is presented. Range of motion is crucial step for quantifying human motor disabilities. An image capturing system is designed and implemented to take images from the hand in vertical and horizontal views. Markers are placed on the hand to enable localization of the hands joints. Using the marked image of the hand, joints positions are extracted and range of motion is calculated via trigonometry relationships. The presented method is reproducible, costefficient and requires less human interaction compared to previous methods. In order to evaluate the accuracy of the method, standard goniometry was considered. Comparison between goniometry readings and the output of our algorithm showed that the presented method is capable to measuring range of motion values accurately.