Image Thresholding on Segmentation of Teeth in Dental Panoramic Radiographs

Panoramic radiographs are usually taken to examine dental diseases. Other works used them for early detecting individuals with general skeletal bone diseases, such as osteoporosis. However most works in computer-aided diagnosis on panoramic radiographs heavily depend on the performance of the segmentation method, due to unevenly illumination and low contrast of the images. We conduct an exhaustive survey for four excellent image thresholding methods and provide their performance comparison on experimental images consisting special area of interest on panoramic radiographs. The methods include Multistage Adaptive Thresholding (MAT), Otsu, Hierarchical Cluster Analysis (HCA), and Fuzzy Sets Type II methods. The comparison is carried out based on the combined performance measures. We identify the MAT method perform better over teeth image applications.