Discovering of cancer is the most extreme interesting investigation space for researcher in the early period.The method used to identify lung cancer with lung CT images in the begining stage so that the survival rate of the patient is high. The methods involve with several stages such as image aquizition,filtering techniques, segmentation techniques, morphological operations and classification.Here in this paper segmentation is perform to disintegrate an lung CT images where gradient magnitude technique is to extract certain essential features from the segmented images ,further extracted features are considered for support vector method to classify whether it is a malignant or begign lung CT images.