基于Openmv和STM32的车内温度预警系统

In⁃car temperature warning system based on Openmv and STM32

随着全球夏季温度升高,婴儿甚至成人在车内被锁而导致晕倒的现象频频出现,基于有效解决该问题从而保证人们生命安全的目的,提出基于Openmv和STM32的车内温度预警系统,利用Openmv控制舵机完成云台360°全方位无死角人脸的追踪,同时利用Haar算子和LBP特征完成人脸识别和监测。倘若车内有目标人物则利用NB-IOTBC26模块和SIM900A模块进行温度检测、数据上传、短信预警、蜂鸣器预警和风扇物理降温。通过人工改变车内实际环境进行试验并完成功能测量和数据记录。实验数据表明,该系统人脸识别误差低于7.5%,精度较传统算法提高了55.06%,总体耗时较传统算法降低了77.77%,有效地保障了车内环境的安全性。

As the global summer temperature rises,babies and even adults are locked in the car and cause fainting frequently.For the purpose of effectively solving this problem and ensuring the safety of people’s lives,an in⁃vehicle warning system based on Openmv and STM32 is used.Openmv controls the steering gear to complete the 360°omni⁃directional tracking of the face without blind spots on the gimbal.At the same time,the Haar operator and the LBP feature are used to complete face recognition and monitoring.If there is a target person in the car,use the NB⁃IOT BC26 module and SIM900A.The module performs temperature detection,data upload,SMS warning,buzzer warning and fan physical cooling.The actual environment in the car was manually changed to test and complete the functional measurement and record data.The data showed that the error of face recognition was less than 7.5%,the accuracy was improved by 55.06%compared with the traditional algorithm,and the overall system time consumption was reduced by 77.77%compared with the traditional algorithm.Effectively guarantee the safety of the car environment.