Abstract

Amongst diverse cancers, cancer of lung be measured the foremost reason of cancer demise including utmost\ndemise pace. Nodules lying on lungs have distinct structure circle otherwise coiled and during various circumstances\ncomposes the recognition complex. In this learning, system has been urbanized for detection before time via images from\nComputerized Tomography (CT) scanner along with distinguishes among malignant and benign tumors. Lung cancer\ndetection process has four steps towards detecting the existence of cancer nodule within lung. Pre-processing phase,\nsegmentation, feature extraction and lung cancer cell classification are the steps. Pre-processing stage includes image\nenhancement. Improved CT image of lung after that conceded all the way through segmentation stage. Commencing the\nsegmented output features be extracted to visualize the occurrence of deformity of lung. The planned scheme gives nodules\nsegmentation lying in particular section with Watershed Transform as well as ensures classification among malignant nodules\nand benign through Artificial Neural Network. The accuracy and sensitivity of 90% and 93.30% is acquired within the\nsystem.