現代社會中將科技結合醫學是一股相當流行的趨勢。運用自動化的電腦技術，我們可以全自動的或是系統化的將一些醫學上的資料加以量化，儲存乃至進一步的對它們作統計與分析。藉由這樣的技術，不僅可以快速且有效率的將醫學用資料量化建檔，並且可以提供醫學人員在直觀的專業判斷之外還有客觀的統計數據資料可輔助參考，甚至可做為對病情做進一步研究探討的依據。
在這篇論文中做的是針對過敏性黑眼圈的偵測與量化。醫師們可藉由觀察患者之黑眼圈的嚴重程度來判斷其過敏症狀的病情嚴重與否。除了一般常用的CCD影像，將還會另外探討觀察患者的紅外線溫度影像來觀測黑眼圈的可行性。我們發展一套電腦演算法，藉著分析患者的CCD影像，自動化的將患者的黑眼圈區域定位出來並對它的面積大小以及深淺程度做量化分析。而紅外線體溫圖由於其眼睛的位置難以觀察並加以定位，因此我們會先將CCD影像以及紅外線體溫影像做完對位之後，利用CCD影像定位好的眼部位置套用在紅外線體溫影像上，再進一步對其眼眶以下的範圍去找溫度較高的那一塊區域對其作偵測及量化，最後再將兩種影像找出的黑眼圈區域的結果做量化並加以比較與討論。
我們的方法可以全自動的在一般影像及紅外線體溫影像上作黑眼圈範圍的定位及量化。並透過分析比較兩者量化的結果，來了解紅外線體溫影像所觀察到的高溫區域和CCD影像所觀察到的黑眼圈區域之間的關係。利用我們提出的一般影像及紅外線體溫影像之間的對位方法，也可以幫助我們將不易判斷五官位置的紅外線影像，用非監督式的方法找出其五官的位置，對於未來應用在臉部的紅外線體溫影像的分析實作上將有所助益。

Nowadays, the combination of computer technology and medical technology is an unstoppable trend. By using the computerized method, we can quantify and store the medical data automatically and systematically. The quantized data is an objective measurement that can be used to help doctors diagnose the symptoms of allergic disease beyond their professional judgment.
In this thesis, we are concerned about how to find a fully automatic method to detect and quantize the region of allergic shiners. Beside experimenting in the visible CCD images we generally used, we will also investigate the possibility of detecting the region of shiners in the infrared thermal images. We developed a method to locate the region of shiner automatically in visible CCD images and then quantize the area and darkness of the detected shiner region. Since the eyes is difficult to locate in IR image, we propose to register the CCD image and IR image first, and then use the location information of eyes which has been detected in CCD image to locate the eyes in IR image. Once we know the location of eyes in IR images, the high-temperature region of the infraorbital area can be located. Finally, we compare the quantized results of the two kinds of images in order to find out the relationship between the region of high-temperature and region of allergic shiner in order to determine whether we can detect a reasonable region of shiner in IR image or not.
In this thesis, we can fully automatically locate the region of shiners in both CCD images and IR images, even with the unstandardized data. By comparing the quantized results, we can find out the relationship between the region of high-temperature observed in IR images and region of allergic shiner observed in CCD images. The method we proposed to register the CCD images and IR images can help us locate the facial features in IR image in an unsupervised way. It will be useful for future works which also apply on IR face images.

[1] Chen CH, Lin YT, Wen CY, et al. Quantitative assessment of allergic shiners in children with allergic rhinitis. J Allergy Clin Immunol. 2009;123:665-671, 671.e1-671.e6.

[2] M. Irani and P. Anandan, “Robust Multi-Sensor Image Alignment,” Proc. DARPA Image Understanding Workshop, vol. 1, pp. 639-647, May 1997.

[3] G. Friedrich, Y. Yeshurun, Seeing people in the dark: face recognition in infrared images, in: Second BMCV, 2003.

[4] Y. T. Pai, S. J. Ruan, M. C. Shie, and Y.C. Liu, "A Simple and Accurate Color Face Detection Algorithm in Complex Background," 2006 IEEE Int'l Conf. on Multimedia and Expo, pp.1545-1548, Jul. 2006.

[5] Zhang SC, Liu ZQ (2005) A robust, real-time ellipse detector. Pattern Recognition 38:273-287

[6] M. Dobes, J. Martineka, D.S.Z. Dobes, and J. Pospisil, "Human Eye Localization Using the Modified Hough Transform," Optik, vol. 117, pp. 468-473, 2006.

[7] A.S. Aguado,M.E. Montiel,M.S. Nixon,On using directional information for parameter space decomposition in ellipse detection, Pattern Recognition 28 (3) (1996)
[8] Eldad Haber & Jan Modersitzki “Multi-Modal Image Registration using Normal Gradient Fields” Emory University, Atlanta, GA University of Lubic, Germany (2005).