| Knee osteoarthritis is the common disease in the elderly clinical, with the ageing of the population, its incidence is rising. Lesions results usually leads to joint pain and dysfunction. At present, although there are many methods in the treatment of osteoarthritis of knee joint, but can not fundamentally solve the influence function on patients with Osteoarthritis outcome brought. Patients with the appropriate master, passive functional exercise, can enhance muscle strength and endurance, promote blood circulation, increased joint range of motion, maintaining joint stability, on the stable condition, relieve symptoms have played a positive role, so sports physical therapy has great value in the clinical application of osteoarthritis. In the training process, if not according to certain criterion of sports, like sports too strenuous or movement is too light, will aggravate the condition of the patient or not up to the effect of training. Therefore, this paper will design a monitoring system based on ZigBee for the osteoarthritis movement therapy, criterion to distinguish the patient’s movement. The main work of this paper are as follows:(1) Analysis the motion model of human, we present a new scheme to obtain the information of lower extremity. The scheme uses multiple three-dimensional acceleration sensor, placed in the particle position lower leg and thigh, using acceleration information acquisition system can effectively obtain the acceleration information of upper limb.(2) Based on the multi particle motion model of the human body, design the acceleration data acquisition system based on ZigBee. Design the circuit of the Zigbee terminal nodes and coordinator nodes According to the requirement.(3) Design the recognition algorithm of knee osteoarthritis common action. The main steps are preprocessing the acceleration signal, extracting the effective action signal, getting the acceleration signal of two identical length by normalized processing, extracting the feature of signals by using the wavelet transform, Then selecting the feature by using of the particle swarm algorithm and getting the strongest feature vector classification ability, Finally using the feature to discriminate the normative of the action.(4)Test the system performance by experimental platform. The classification results show; On the stage of leg lifting, keeping and legs stage the four kinds of action (standard actions, curved legs, leg and jitter) recognition rate were above 95.9 ± 1.2%.This paper design a movement therapy monitoring system of osteoarthritis based on ZigBee, the system can effectively discriminate the non normative action, satisfy the requirements of monitoring system. The high performance price ratio, small volume, low power consumption, it have practical application value in the movement therapy monitoring field of osteoarthritis exercises. |