基于Harris算法的红外人脸识别测温系统设计

Design of Infrared Face Recognition Temperature Measurement System Based on Harris Algorithm

本文指出人的体温是反映人体健康状况的重要指标,体温异常与某些疾病有着密切的联系;从防疫筛查的角度出发,公共场合人流体温检测已成为筛查潜在病毒感染者的重要手段。为此,本文设计了一套基与红外测温和人脸识别技术的门禁系统。系统采用STM32F103C8T6单片机作为主控模块,以OpenMv4 H7作为摄像头采集模块,采集人脸图片,采用Harris特征提取算法进行特征点提取,实现人脸学习识别功能;以MLX90614红外测温传感器测量人体或者物体温度,并与设定的阈值比对,超过阈值实现语音与灯光报警。经测试,本系统对人体以及物体温度测量的误差控制在0.5℃之内,人脸识别的准确率达到98%,系统操作简单,运行可靠,较好地实现了功能。

This paper points out that human body temperature is an important indicator to reflect the health of human body.Therefore,in this paper,a set of access control system based on infrared detection and face recognition technology is designed.In this system,STM32F103C8T6 microcontroller is used as the main control module,and OpenMv4 H7 is used as the camera acquisition module to collect face pictures,and Harris feature extraction algorithm is used to extract feature points to realize the function of face learning and recognition;MLX90614 infrared temperature sensor is used to measure the temperature of the human body or object,and the temperature is compared with the threshold,and the voice and light alarm can be realized when the threshold is exceeded.Through the test,the system to the human body and object temperature measurement error control is within 0.5℃,the accuracy of face recognition is 98%,the system is simple to operate,reliable to run,and the function is better realized.