透過您的圖書館登入
IP:52.15.59.163
  • 學位論文

低解析度熱感應器之熱源定位精度與校正研究

Calibration of low-resolution thermal sensors and enhanced heat source positioning accuracy

指導教授 : 施吉昇

摘要


世界各國開始進入老齡化社會,長期照護開始成為重要的行業。在長期照護系統中,常用普通的RGB攝影機進行記錄,這會導致老年人的隱私問題。部分機構改用可以不分晝夜進行紀錄的紅外線熱感應器,但如果使用解析度較高的型號,仍然會有隱私問題。因此,我們使用超低解析度紅外線熱感應器來記錄老年人的身體狀況。 在這項研究中,我們使用了具有相機校正功能的單個超低解析度紅外線熱感應器來檢測熱源的移動距離。 一般的紅外線熱感應器是可以使用自熱式校正板來做相機校正,但是此研究中使用的紅外線熱感應器解析度遠低於一般市面上常見的機型。因此我們在固定尺寸的感測平面上記錄熱源的移動範圍,並結合熱感應器上顯示的相應位置以製作模擬校正板。通過這個模擬校正板,我們可以對超低解析度的熱影像做校正。 我們使用相機校正來獲得紅外線熱感應器的相機矩陣,用來還原感應器中的每個熱電堆感測到的實際場域大小。有了這個正確的感測場域,可以讓我們用來計算熱源的移動距離。我們的貢獻是如何製作模擬校正板,並使用校正後的圖像計算熱源的移動距離。使用我們的方法可將熱源移動的測量距離誤差降低至 6.9%。

關鍵字

紅外線影像 相機校正 IoT

並列摘要


The world has begun to enter an aging society, and long-term care has become an important industry. In the long-term care system, ordinary cameras are often used for recording, causing privacy problems for the elderly. Some of them begin to use infrared sensors, which can record day and night while using a higher resolution model, and it still causes privacy problems. To solve this privacy problem, we choose thermal sensors to record the activities of the elderly. In this study, we used a single ultra-low-resolution thermal sensor with camera calibration to detect the moving distance of the heat source. Since the resolution of our thermal sensor is too low, it is impossible to use a self-heating calibration board. Therefore, we record the moving range of the heat source on a fixed-size sensing plane and combine the corresponding position shown on the sensor to make a simulated checkerboard. We can calibrate the ultra-low-resolution thermal image through this new checkerboard. We can use the camera calibration to get the camera matrix to restore the actual size of the sensing field in the corresponding grid of the sensor. Use this actual size, and it can help us calculate the moving distance of the heat source. Our contribution is how to make a simulated calibration board and use the corrected image to calculate the moving distance of the heat source. Using our method can reduce the error of the moving distance to 6.9\%.

並列關鍵字

thermal imaging camera calibration IoT

參考文獻


[1] “Mathworks distortion example.” [Online]. Available: https://ww2.mathworks.cn/ help/vision/ref/cameraintrinsics.html
[2] “Grideye.” [Online]. Available: https://na.industrial.panasonic. com/products/sensors/sensors-automotive-industrial-applications/lineup/ grid-eye-infrared-array-sensor
[3] “Raspberry pi 3 model b+.” [Online]. Available: https://www.raspberrypi.org/ products/raspberry-pi-3-model-b-plus/
[4] “Arduino mini.” [Online]. Available: https://store.arduino.cc/usa/arduino-pro-mini
[5] “Infrared array sensor grid-eye (amg88) datasheet.” [Online]. Available: https: //datasheet.octopart.com/AMG8833-Panasonic-datasheet-62338626.pdf

延伸閱讀