Abstract: Recognized for its severity and widespread impact, cancer remains a leading cause of death worldwide, claiming countless lives annually. Touching upon cancer genres, pleura tumor tops the list as the most widespread and deadliest form. In the detection of pleura tumor, CT scans (computed topography) play a crucial role by offering a detailed view of tumor presence and progression. Although CT scans function as primary choice pleura tumor determination, the subjective nature of visual analysis can introduce inaccuracies and obstacles. Thus, the use of image processing methods has become prevalent in medaical research, particularly for the early determination of pleura tumor. This paper presents a new methodology for determine pleura cancer from given inputs. This study presents an algorithm aimed at detecting pleura tumor, which incorporates methods such as median filtering to preprocess images and a set of techniques used in image processing and computer vision for analyzing and manipulating shapes within images are employed to divide the bronchial area of focus. Geometrical features derived from this segmented area are then utilized in a support vector machine classifier to discern between healthy and peculiar CT scan images