罹患新冠肺炎產生的主要症狀之一為發燒，因應此特性去加以防止病毒傳播，公共場域均開始設置溫度管控機制，過往確認發燒病患，多採用額溫槍體溫量測，不僅耗時且耗力，也可能因近距離接觸而產生感染風險，少部分中大型場域及公司機關使用紅外線熱像儀大範圍測溫，然而常因其他非人臉熱源干擾，產生錯誤誤告警示，且人潮流量大時，監督人員可能無法即時兼顧。 本文介紹一些紅外線熱影像的觀念與原理，透過硬體式的熱像儀，非接觸即可測量人體臉部溫度，加入軟體研究「多人臉動態偵測」及「溫度補償校正」兩項功能，將硬體透過軟體優化，有效提高偵測準確率，整合而成進階版的智能版熱感系統。 實際架設系統進行數據分析三個驗證，第一是人臉溫度偵測成功率大於95%，第二是非人臉溫度誤偵測率小於5%以及第三是距離與溫度標準差σ小於0.5℃；有著實驗成果可精準擷取十個以上動態人員體溫判讀，透過告警功能通報主動限制體溫異常者進入，快速消化排隊人流，達到智能防疫的效果。 ;One of the main symptom of COVID-19 is fever. In response to this feature, to prevent the spread of the virus, temperature control mechanisms have been set up in public places. In the past, patients with fever have been confirmed to use a forehead thermometer to measure their body temperature. This is not only time-consuming and labor-intensive, but also there may be a risk of infection due to close contact. Some medium and large fields and corporate agencies use infrared thermal imaging cameras to measure temperature in a wide range. However, it is often caused false alarms by interference from other non-face heat sources, and when there are too many people, the supervisor may not be able to take care of it immediately. This article introduces some concepts and principles of infrared thermal imaging, using a hardware thermal imaging lens to non-contact measurement of human face temperature, adding software research for the two functions of "Multi-face motion detection" and "Temperature compensation correction". Optimize the hardware through software to effectively improve the detection accuracy and integrate it into an advanced version of the intelligent thermal system. Practical setup of the system for data analysis to verify three parameters. The first is that the face detection success rate is greater than 95%, the second is that the non-face temperature false detection rate is less than 5%, and the third is that the standard deviation of distance and temperature σ is less than 0.5°C. The experimental results can accurately capture the body temperature of more than ten dynamic personnel, and actively restrict the entry of people with abnormal body temperature through the alarm function, which could also quickly digest the flow of people in line, and achieve the effect of intelligent epidemic prevention.
