Current Volume 6
This paper proposes an automated target classification algorithm using Soft Computing Technique to extract feature vectors for craniofacial features in cephalometric radiograph. The proposed work is based on image segmentation and classification technique, using Top-Hat filter and soft computing technique. The preprocessing includes Histogram Equalization [HE] to improve the contrast level. The preprocessed image is then segmented for image analysis. Segmentation will extracts small elements and details. The features are extracted by subjecting the radiography to the Gray Level Co-occurrence Matrix [GLCM] algorithm. The resultants are the frontend of support vector machine. Vectors, which posse’s landmarks, are separated from all other vectors. The centroid points, automatically determined from GLCM feature vectors, are the location of landmarks. The landmark points which are serving as a guide for construction and measurement of planes. The resultant segmentation, are the frontend of Convolutional Neural Network [CNN], classifies the normal cephalometric dental image into abnormal cephalometric dental image. According to the performance evaluation of SVM, accuracy and sensitivity of the classification were automatically estimated using True Positive [TP], False Positive [FP], False Negative [FN], and True Negative [TN]. The accuracy of CNN is, number of correct prediction divided by total number of prediction. Finally results of the classification, which has 92.6% accuracy for Non linear kernel and 70% accuracy for linear kernel of SVM and CNN classifier, gives 78.01% accuracy. The performance measurement of soft computing, are used to evaluate the dento-facial relationship, study of growth and development, and also for treatment planning.
Ramya K , Dr. A. Banumathi "Performance Analysis of Dental Deformities in Cephalometry Image using Soft Computing Technique" Iconic Research And Engineering Journals Volume 4 Issue 6 2020 Page 46-51
Ramya K , Dr. A. Banumathi "Performance Analysis of Dental Deformities in Cephalometry Image using Soft Computing Technique" Iconic Research And Engineering Journals, 4(6)