Abstract: Porcelain tooth is an ideal prosthesis of oral diseases. Limited by the fabrication process, there are always some internal defects in porcelain teeth, such as stomas, shrinkages and bubbles. These defects affect the service life and strength of porcelain teeth. It is proposed to scan the porcelain teeth by MicroCT. Then the defects can be displayed clearly through image processing. Subsequently, some features of defect are extracted, including areas and positions. Finally, the statistics of these features are analyzed in order to giveadviceto improve the fabrication process of porcelain teeth. The analysis results show that the distribution of area of internal defect obeys Weibull distribution, that is, the number of internal defects with large area is small but the number of internal defects with small areas is quite large. It reveals that most of defects are caused by porosity and shrinkage, which often result in small area defects. Additionally, the defects are mainly located at crown and joint surface, where the stress concentration is often located. Therefore, there are two methods to reduce the defects. One is reduce the stress concentration and the otherto improve the gas exit.

Key words: MicroCT, image process, internal flaws of porcelain, statistical analysis