Globally, lung cancer is a prevalent reason for death for individuals. The most common form of lung cancer has a high death rate. If lung cancer is identify and care for in its early on stages, screening can assist identify the disease at an earlier stage. Therefore, a patient's chances of survival are increased by early identification of lung cancer. Lung computed tomography (CT) scans might represent a more reliable method of identifying cancer than MRIs and X-rays. DICOM (Digital Imaging and Communications in Medicine) is used to scan the lungs and obtain images. We improve and smooth photos using a variety of preprocessing approaches. Edge detection and thresholding (ROI) are used in this work to section the lung tumor's region of interest. Lastly, we use Support Vector Machine to compute a number of geometrical aspects of the retrieved ROI and categorize the results into both benign and malignant degrees of severity (SVM). Data is classified using supervised machine learning techniques called SVM classifiers. One of their advantages is their ability to handle small amounts of high dimensional data. Every characteristic is used as an input to evaluate the SVM's efficiency. As such, lung cancer in CT scans is diagnosed using a method that uses methods for processing images.