Posture Recognition Analysis And System Design Based On Plantar Pressure

Distinguishing walking,standing and sitting postures based on plantar pressure can help formulate long-time work posture adjustment strategies and reduce health risks in the workplace.In order to improve the ability of posture recognition of plantar pressure,this thesis proposes a new posture recognition index based on the pressure ratio between the two anatomical plantar regions.The pressure ratio contains the information of motion pattern and motion inertia of posture,and is less affected by individual weight,so it is more suitable to posture recognition.Furthermore,a plantar pressure posture recognition system with the ability to recognize walking,standing and sitting postures was designed and manufactured,and its effeciency was verified by experiments.First,in order to determine the plantar pressure index for posture recognition,the plantar pressures of 30 participants during walking,standing and sitting were collected by the F-SCAN plantar pressure measurement system,and the plantar area was divided into 7 subareas according to its anatomical structure.According to the distribution of plantar pressure in 7 subareas,it was found that the sitting posture can be distinguished from standing and walking postures via its characteristics of zero plantar pressure in the forefoot subarea.Further,the two-factor repeated measurement analysis of variance was used to analyze the differences of different identification indicators(peak contact pressure,average contact pressure,and pressure ratio)in walking and standing postures.The d size effect was used to quantify the difference,and it was found that the effective recognition of walking and standing can be achieved through the pressure ratio of two palantar subareas.At the same time,the value ranges of the recognition index corresponding to different postures were given.Then,according to the value ranges of the recognition index,a wearable plantar pressure posture recognition system was designed.The whole system is divided into lower computer and upper computer.The hardware part of the lower computer uses STM32 as the control core,the FSR-402 piezoresistive sensors as the plantar pressure measurement elements,the voltage conditioning amplifier circuit as the data processing module,and the Bluetooth as the wireless data communication module.The hardware part of the upper computer uses a PC.In the software part,the lower computer completes voltage acquisition,analog-to-digital conversion and data transmission.The Lab VIEW program in the upper computer receives the data,converts the sensor voltage value into the pressure value,and calculates the recognition index value for the current posture to complete the posture identification.Finally,the effect of the fabricated plantar pressure posture recognition system was verified by experiments on tiles,wooden boards and foam floors with different textures.Experimental results show that walking and standing postures can be accurately identified,but sitting posture may be recognized as walking posture on a hard surface.The posture error rate of the system is 3.3% on the tile and wood floor,and the system can accurately identify the postures on the foam floor.Compared with the 88% detection accuracy rate of other similar systems,the overall detection accuracy rate of this system is 96.7%.It has good recognition performance for walking,standing and sitting posture,and is expected to be applied to the workplace to help workers change their postures in a timely manner and reduce health risks.