Research On Medical Infrared Thermometer And The Key Technology

Body temperature is one of the basic vital signs of the human body,which is an important indicator reflecting the healthy life of the body.At present,the higher accuracy of the instruments for measuring temperature directly affects the diagnosis of the disease,the observation of the condition and the development of the treatment plan.However,traditional mercury thermometers and electronic thermometers obviously cannot meet the requirements of rapid measurement.What's worse is that this kind of contact measurement has the risk of cross infection.Recently,the infrared thermometers as the focus of public concern mainly results from its non-contact,short measurement time and high sensitivity.The infrared ear thermometer,the infrared forehead thermometer and the infrared thermograph are common in our daily life.When using the infrared ear thermometer,the sensor should be aligned with the eardrum.If not properly handled,the temperature measured will not be accurate;The infrared thermometer is affected by the ambient temperature and distance,yet,the current accuracy is not too high;Infrared thermal imagers have high temperature measurement accuracy,but they are expensive.Aimed at improving the accuracy of the experiment needs of measuring temperature,in this paper,we design a medical infrared thermometer based on BP neural network distance compensation when the cost is controlled at a certain level.First of all,this paper carried out the overall design of the medical infrared thermometer.It mainly includes the design of hardware circuits and software programs.Hardware part:According to the functions that each sub-module needs to implement,complete the selection of related devices,and design its related interface circuit and peripheral circuit.Select the related type of device according to the function of each sub-module.Furthermore,we design related interface circuits and peripheral circuits.By comparing several infrared temperature sensors,we chose a thermopile type infrared temperature sensor with a relatively easy signal processing circuit,good stability,and low cost.Software part:Write the program to realize MLX90614 measuring infrared temperature,DS18B20 measuring ambient temperature,HC-SR04 measuring distance,LCD real-time display and other functions.In addition,we have studied the primary technology of temperature compensation algorithm based on BP neural network for medical infrared thermometer.By comparing several temperature compensation algorithms,BP neural network algorithm with strong fault tolerance and strong generalization ability is selected to measure distance and ambient temperature compensation.Experiment were carried out to verify the practical effectiveness of our algorithms,and the results showed that,the error was within±0.1~o C.It can effectively reduce the influence of environment temperature,distance and other factors on temperature measurement accuracy.