透過您的圖書館登入
IP:3.138.114.132
  • 學位論文

Recognition of Down Syndrome Based on Active Appearance Model

基於主動外觀模型之唐氏症辨識

指導教授 : 陳永昌
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


If the chromosome is unusual, it will cause the distortion on the facial features or the body. It is common that most of patients of Down syndrome have a variety of multiple characteristics. Sometimes, a patient has some obvious distinguishable characteristics, and some characteristics are indistinguishable. So, we have to find characteristics and rules to recognize the patients of Down syndrome. We choose the idea of face recognition to construct a disease recognition system by active appearance model (AAM) and support vector machines (SVM). Experimental results show that the accuracy of our system is 88.9% to recognize Down syndrome case correctly.

並列摘要


依目前醫學對於罕見疾病患者的診斷,最準確的檢驗是採用染色體,但染色體檢驗必須花費昂貴的金錢以及較長的時間去做DNA 分析。 因此這篇論文主要目的是希望建立一個唐氏症疾病辨識系統,且由人臉辨識的角度來看待這個問題。所以只需要擁有人臉資訊或明顯特徵,即可來幫助醫生做簡單、快速疾病診斷。 唐氏症疾病患者可能因某些染色體上的基因異常,導致五官或是身體上的畸變,我們希望可藉由臉部的明顯特徵來做唐氏症辨識。先前實驗室學長的研究方法以量測為主體。在這篇論文我們將從另一個角度(人臉辨識)來看待這方面的問題,以讓疾病辨識系統有個基本雛型,目前我們先將問題縮小在幼童唐氏症疾病的辨識。 唐氏症疾病患者五官有下列幾個明顯特徵,例如 : 五官比例特殊、鼻子太扁、耳朵外觀怪異、臉頰過於肥厚扁平。在這個唐氏症辨識系統裡,我們只針對正面的人臉影像,首先利用AAM 來做人臉特徵點萃取,可以有效的減少搜尋特徵點的時間。接下來,將由AAM 所得到的外觀模型作正規化,找個相同基準點以後建立分類器的模型。最後,使用SVM 當我們的分類系統,比較先前正規化後的外觀模型之間差異。實驗結果的顯示,在做唐氏症患者的辨識上,我們具有九成左右的成功辨識率。

參考文獻


[3] Xiaoguang Lu, "Image Analysis for Face Recogintion," Dept. of Computer Science & Engineering Michigan State University, East Lansing, MI.
[6] G. Yang and T. S. Huang, "Human Face Detection in Complex Background", Pattern Recognition, vol. 27, no. 1, pp. 53-63, 1994
[8] L.M. Zhang and P. Lenders, "Knowledge-based eye detection for human face recognition", Proc. Knowledge-Based Intelligent Engineering Systems and Allied Technologies, vol.1 pp. 117~120, 2000
[9] D. Reisfeld and Y. Yeshurun, "Robust detection of facial features by generalized symmetry", Proc. International Conference on Pattern Recognition, pp.117~120, 1992
[11] E. Osuna, R. Freund, F. Girosi, "Training Support Vector Machines: An Application to Face Detection", Proc. IEEE Conference on Computer Vision and Pattern Recognition, pp. 130-136, 1997

延伸閱讀