牙齒X光片與牙齒治療之分類方法

Abstract

身分辨識技術近年來廣泛的被使用在日常生活中，例如：虹膜、指紋、聲紋、臉部辨識…等。各種生物特徵迅速運用於門禁控管、交易安全、法律執行等相關領域中。但是在法律界領域中，罹難者的行為特性以及生物特徵往往因為重大事故或是身體組織受到嚴重損害。牙齒是人體中最堅硬的身體組織，可以抵抗高溫以及強烈撞擊，非常適合作為罹難者之辨識之依據。由於傳統牙齒辨識方法須以人工辨識，對龐大的資料庫進行人工比對是曠日費時且效果不佳的作法。因此，有人提出自動化牙齒身分辨識系統，利用罹難者(post-mortem)的牙齒X光片影像和資料庫中所記錄的生前(ante-mortem)牙齒X光片進行比對，藉此找到最吻合的人，達到身分辨識之目的。此外，牙齒中各種類型之病灶處理，包含填補、牙套、植牙…等。每一種牙齒治療具有不同特性，也可以作為輔助身份辨識之依據。所以本論文中提出了兩個方法，一種是階層式的特徵萃取法，可以有效的分類牙齒X光片。一種是型態學的特徵萃取法，可以有效的分類牙齒治療形狀。因為每種影像有每種影像不同的處理方式，對這些影像做完分類之後，就可以針對它們的特性進行分析，而達到自動化牙齒身分辨識的功能。

In recent years, identity recognition technology is widely used in daily life, such as: iris, fingerprints, voice prints, facial recognition ... and so on. Varieties of biological characteristics are quickly applied in control, transaction security, law enforcement and other related fields. However, in the field of the legal profession, victims and biological characteristics of the behavior are often damaged cause of major accidents. Teeth are the hardest body tissue that can resist high temperatures and strong impact. It is very suitable as a basis for identification of the victims. Because of traditional method of dental identification required to identify on a large database manually, it is a time consuming and ineffective approach. An automated dental identification system is proposed to find the best fit people by comparing dental radiographs of victims (post-mortem) and the dental radiographs on database are recorded before his death (ante-mortem). In addition, all kinds of dental work, including filling, crowns, implants ... and so on have different characteristic, it can be used as a basis for supporting the identity recognition. Therefore, this paper proposed two methods, one is a hierarchical feature extraction method of classifying dental radiographs effectively. The other is a morphological feature extraction of classifying dental work effectively. An automated dental identification system can be achieved by these effective classifications.