Segmentation of Tooth and Pulp from Dental Radiographs [ ]


Teeth and pulp segmentation for periapical radiographs is one of the most critical tasks for effective segmentation in periapical radiographs, and the process is difficult due to noise, low contrast, and uneven illumination of the dental radiographs. For accurate segmentation, images must be preprocessed either by image enhancement or image transformation or both so that the aforementioned problems will be reduced as much as possible. In this work, proposed an effective scheme to segment each tooth and pulp in periapical radiographs. The method consists of four stages: image enhancement using Contrast Limited Adequate Histogram Equalization(CLAHE), local singularity analysis using Hölder exponent, connected component analysis, and tooth segmentation using Chan-Vese segmentation.