Study On Hand Virtual Rehabilitation System Based On Hand Gesture Recognition

Hand is one of the important organs of mankind,hand injury will seriously affect people's normal life,hand trauma,cramps and stroke is the main cause of hand dysfunction.The current recovery of the function of the opponent mainly depends on the training or massage of professional medical staff.Studies have shown that timely and effective rehabilitation training on patients with hand recovery has a very important impact.However,the existing hand function rehabilitation system are more boring,lack of fun,which has brought negative emotions for the patient in the rehabilitation process and is not conducive to the rehabilitation of patients.In order to carry out professional,step-by-step rehabilitation training for patients with hand dysfunction,and to increase the patient's interest and training time in rehabilitation training and safety training.In this paper,the hand virtual recognition system based on gesture recognition is designed under the premise of comprehensive fun,interactivity and cheapness.It mainly contains the following aspects:First of all,we designed a low-cost portable virtual rehabilitation test box,and completed a variety of hardware design,human-computer interaction interface and visual design.Through the interaction between the patient and the game scene complete the recognition of rehabilitation gesture action,while enhancing the fun to achieve a good hand interaction and rehabilitation effect.Secondly,a rehabilitation gesture recognition algorithm is proposed.The algorithm first extracts the external contours and feature points of the gestures,and then describes the feature vectors of the feature points based on the shape context algorithm.At last,the Euclidean distance is used as a measure factor to identify the similarity between gestures.This paper defined seven kinds of rehabilitation gestures for patients with functional dysfunction,and established a database of rehabilitation gestures.Compared with the previous simple gesture recognition,a more targeted rehabilitation gesture recognition scheme and gesture evaluation function were put forward.Finally,a dynamic recovery gesture recognition algorithm is proposed.Using the CTtracking algorithm to track the hand area,and then preprocessing the hand information to obtain the maximum connected area of the hand in the tracking box.By obtaining the smallest circumscribed rectangle of the hand connected area,the center of the smallest circumscribed rectangle is approximately the center of gravity of the hand.At last,the shape of the hand and the trajectory of the motion trajectory combined into the feature vector to complete the rehabilitation action recognition.