A Reliable and Fast Detection Technique for Lung Cancer Using Digital Image Processing

Abstract

This paper depicts a computer-based algorithm that analyzes different normal and cancer cell affected images. Several image segmentation and enhancement algorithms have been used in this proposed system—many images collected from a local hospital and from open access internet database. Histogram equalization has been used in the end stage and compared with the original normal and affected cancer cell image curve. The novelty and uniqueness of this paper are that the output curve of the proposed system has to match with two sets of algorithms (AHE, watershed segmentation) to ensure lung cancer. If both two curves are matched similarly concerning the original images we collect, a conclusion can be drawn. Results were 82% same after using a variety of lung images.