Research On Sitting Posture Recognition System Based On Multi-sensor

With the continuous development of science and technology,people's living standards are getting higher and higher.They more and more concerned about health.Sitting posture,as the most common state of life,the impact on personal health causes a lot of researchers' highly concern.At present,the sitting recognition technology is mainly based on in two aspects.There are visual technology and sensor-based technology.But for wheelchair users,the process of moving the light received a greater impact,and the use of cost considerations.This paper selects the sensor recognition technology.A single type of sensor has the problem of low recognition rate and poor stability.To solve these problems,design a simple operation,identify the high accuracy of the posture recognition program and system is of great significance.According to the characteristics of the acceleration sensor and the pressure sensor,it can provide users with a more effective basis.Last this function is achieved with EMG for fatigue monitoring.The corresponding recognition algorithm is designed for the aspects of sitting and sedentary reminder.The main work and achievements of this paper are as follows:(1)The design of the body posture recognition system including the acceleration sensor identification module,the pressure sensor identification module and the EMG sensor identification module is completed.The module has been completed from the selection of components,layout,to the final circuit design.Established a simple operation and good stability sitting position recognition.(2)A threshold-based recognition algorithm is designed for the acceleration sensor module.The appropriate threshold is selected by analyzing the experimental data,and the threshold is corrected by using the root mean square as the eigenvalue to solve the problem that the threshold is not common due to the different habits of different people,So that recognition is more common.(3)This thesis designed the body posture classification algorithm.The algorithm of fuzzy theory and D-S evidence is designed to identify the human sitting position,improving the accuracy and stability of the sitting posture solution,solving the problem of low recognition rate and poor stability of single sensor,and promote the application of posture recognition technology Finally,the experimental results are verified by experiments,and the reliability and accuracy of the algorithm are verified.