非制冷红外热像仪人体表面温度场测量及误差修正

利用非制冷红外热像仪测量人体表面温度场，除具有快速、非接触和量程短等特点外，还对测温精度有较高要求。针对非制冷微测辐射热计热像仪测量精度受环境、机芯温度影响较大的问题，提出一种对热像仪使用温度与标定温度之差引起的测量误差进行修正的方法。即对分别测得的环境温度、机芯温度和灰度两组数据，由支持向量机拟合得到环境温度和机芯温度误差修正模型；实际测量时，分别由热电偶和置于热像仪中的传感器测得环境温度和机芯温度后，根据误差修正模型对环境和机芯温度变化引起的热像仪测量误差进行修正，获得较为准确的人体表面温度场数据。实验结果表明:该修正方法，与经标定的高精度热电偶测温相比，可使测量距离2 m时的测温误差减小50%。

Temperature field of body surface measurement by uncooled infrared thermal imager and error correction

Measure body surface temperature with uncooled infrared thermal imager has many features such as fast, non-contact, short scope of measurement and higher accuracy requirement. But temperature of core and environment will interfere the accuracy of micro-bolometer to a great degree. A new method was proposed to correct temperature error caused by the difference of temperature between calibration and measurement. The data of core temperature, environment temperature and grayscale were measured and recorded, and then error correction parameter was calculated by using SVM. In the course of measurement, the environment temperature and core temperature were obtained with thermocouple and sensor in micro-bolometer, and then error correction of environment temperature and core temperature were finished by error correction parameter. By this way more accurate data of human body surface temperature field were gained. Experimental results shows that the error of new corrected method can be decreased at least 50% when the distance of measurement is about 2m, compared with thermocouple of high precision which had been calibrated.