基于MPU6050和神经网络的颈椎病预防系统设计

Design of Cervical Spondylosis Prevention System Basedon MPU6050 and Neural Network

针对目前坐姿识别方案存在的监测数据单一、实时性差等问题,设计了一款预防颈椎病的智能监测系统.该系统以STM32F103C8T6芯片为核心,通过MPU6050加速度传感器与HC-SR04超声波测距传感器相结合的方式实现头颈部信息的采集,利用蓝牙模块将数据信息传送给手机移动终端,手机端使用基于全连接神经网络的坐姿分类模型对用户坐姿的数量及持续时间进行识别统计.经测试,该系统对坐姿识别的准确率达到98.24%,可以满足坐姿快速精确识别的需求,可适用于颈椎、近视等疾病的早期预防.

Aiming at the problems of single monitoring data and poor real-time performance in the current sitting posture recognition scheme,an intelligent monitoring system for preventing cervical spondylosis is designed.The system takes STM32F103C8T6 chip as the core,realizes the collection of head and neck information through the combination of MPU6050 acceleration sensor and HC-SR04 ultrasonic ranging sensor.The bluetooth module is used to transmit the data information to the mobile terminal of the mobile phone.The mobile terminal uses the sitting posture classification model based on the fully connected neural network to identify and count the number and duration of the user’s sitting posture.After testing,the accuracy of the system for sitting posture recognition can reach 98.24%,which can meet the needs of fast and accurate recognition of sitting posture,and is suitable for the early prevention of diseases such as lumbar disease and myopia.