Objectives: This paper develops an algorithm to generate automatically, a feature vector using various statistical texture features of trabecular mandible from digital Orthopantomogram (OPG). These statistical features are computed using Gray Level Co-occurrence Matrices (GLCM) of mandible structure. Analysis: For this paper, region of interest is trabecular mandible bone present in lower jaw. Mandible Segmented from OPG is taken as input to the algorithm. GLCM of this mandible is calculated at

more »

... four angles viz., 0 0 , 45 0 , 90 0 and 135 0 and for two gray scale intensities viz., 128 and 256. Various texture features are calculated from each of these GLCMs. These statistical features are compared and analyzed. Findings: 100 digital OPGs are used for analysis of statistical texture features. Four texture features contrast, viz., correlation, energy, homogeneity are calculated for the four angles mentioned above. Improvement: Texture analysis is performed on various types of images since last 50 years. Locating systemic disease features (e.g., Osteoporosis) from OPG is an upcoming research area. In this area, it is observed that comparatively less literature is available which uses GLCM texture features. The algorithm proposed here devices a texture feature vector for mandible. Analysis of the features is also presented in this paper.