Lung tumor area recognition and classification using EK-mean clustering and SVM

Abstract

This article presents to detect lung tumor, classification and area recognition system. Nowadays, Lung tumor is major cause of death for all the people. Early detection of lung tumor plays an important part to enhance chance for survive to live. Early detecting of tumor is a important role for the treatment where Computed Tomography (CT) screening are consider as appropriate method for detecting the lung tumor. There are five steps involved, they are pre-processing, feature extraction of lung tumor region from CT, image segmentation of lung tumor region, feature extraction, classification of lung tumor as malignant or not. In proposed method involves preprocessing, number of mask can be calculated by thresholding approach, removal of unwanted information from an image. Followed by, suspected Region Of Interest (ROI) can be calculated. For extraction of features, GLCM used for extraction of that suspected ROI. Then, Support vector machine (SVM) classifier classifies those extracted images.