Lung Cancer Detection using Regional and Marker based Watershed Segmentation

Lung cancer proves to be a harmful threat to people who are more commonly used in people who smoke. Of 100 different types of cancer observed in the human body, this is the third largest cancer found with less survival rate. Early detection of lung cancer can increase the survival among humans. In the traditional method they use to identify lung cancer using Active Contour Segmentation Model. First, both lungs are segmented by the active contour model for their initialization, a masking technique is used. This mask is partly preserved by our morphological processes after refinement. The watershed transformation is widely used in many areas of image processing, eliminating the need for any kind of contour connection. The overall performance of the Active Contour Segmentation method is 86%, but we can achieve up to 98% performance of the result can be obtained in our novel Regional and marker based watershed segmentation.