Detection and Classification of Tumor using SVM and ANN with GLCM features in CBIR

Main Article Content

Madhu, Raman Kumar

Abstract

Proposed research has focused on content-based image retrieval in medical field where the tumor is detected from the given dataset. The objective of the research is to reduce the tumor detection time.  This research is considering canny based edge detection, GLCM, ANN, SVM techniques to achieve the objective. The objective of research is to study and analyze various medical image retrieval techniques and propose hybrid technique using feature extraction, feature selection, and classification techniques. Then the performance of proposed technique is evaluated and validated. In this research work, the proposed method comprises four stages. Initially, pre-processing of the images is done. During this stage, the image is resized and RGB to gray conversion is applied. The edge detection mechanism is applied afterward. In the second phase, the image features would be extracted using GLCM, then matching of images with the database is done. Then the tumor is detected according to the features extracted from the image. The third phase is to perform classification using an artificial neural network to detect the tumor and normal image. Then SVM would be applied to detect the shape of the tumor. Finally, the results of the proposed work would be compared to previous research for evaluation.

Article Details

Section
Articles