Document Type : Primary Research paper
Assistant Professor, CSE Department, Bannari Amman Institute of Technology, India
Age is most important factor for every individual in their life. Dental Age (DA) estimation is used for criminal, civil, anthropologic and forensic purposes. Numerous techniques have been provided to evaluate chronological age for these applications. It includes somatic growth measurements which depend on the progress of tooth. Estimation of age has been consumed for long time by the incremental development of tooth. Therefore, DA is measured as substantial fact for establishing individuals’ age. Here, an Elman Neural Network (ENN) with Guaranteed Convergence Particle Swarm Optimization (GCPSO) algorithm is proposed for dental age classification. Initially, OPG input tooth image is preprocessed using Anisotropic Diffusion filter (ADF) for smoothing the image and removal of noise. Teeth image has been segmented by kernel based Fuzzy C-means clustering with Dragonfly optimization then morphological post processing is utilized to enhance the image feature precision. Exact features are extracted and age is classified with ENN-GCPSO. Investigational results determines that ENN-GCPSO obtains better accuracy of 89%, specificity of 83.5%, precision of 72.15%, recall of 92.3% f-measure of 74.54% than the existing classifiers such as ENN-DO, MELM-SRC,RBFN and ANFIS schemes.