Classification and Numbering on Posterior Dental Radiography using Histogram Intersection (情報論的学習理論と機械学習・第15回情報論的学習理論ワークショップ)

概要

論文の詳細を見る
Classification and numbering on posterior dental radiography are useful for forensic and biomedical application. This paper proposed a new method that simple but robust for classification and numbering on posterior dental radiography using histogram intersection. In this method, we utilized the different characteristic between molar and premolar. Molar has wide shape and two roots while premolar has slander shape and one root. We computed the distance between centroid and boundary in each tooth. 36 points in the boundary were extracted. A histogram was created from the distance between centroid and those points. Then similarity among the histograms was computed. From experiment, every tooth has been assigned according to universal dental numbering and classified as their sequence order. Our system achieved average classification precision of 90%.
一般社団法人電子情報通信学会の論文
2012-10-31