Now a days, The digital image recognition techniques are extensively used in a couple of therapeutic domains for image improvement in earlier division and treatment stages, where the time factor is basic to discover the variety from the standard issues in target pictures, especially in various threatening development tumors, for instance, and lung infection. Lung cancer diagnosis is the most important early-stage research field for researcher. The proposed system is designed for the premature detection of lung cancer in two stages. The proposed method consists of several steps such as image acquisition, preprocessing, trivialization, thresholding, segmentation, extraction of features, and detection of neural networks. At first, the lung CT images they're input into the system and then passed through the preprocessing phase of the image using some processing techniques. Lung cancer growth is the most significant sickness cause high death rate. Also, computer vision methods helped determination can be valuable for doctors to precisely recognize the malignant growth cells. Numerous computer vision supported strategies have been contemplated and applied utilizing image recognition and deep learning. To minimize their effect in classification, the JSRT data set is considered to be the most commonly used reference data set to perform experiments. The right segmentation of the lung tumor from X-rays, CT-scans or MRI are the steps towards an integrated diagnostic device for the detection of lung cancer To train this neural network using volumes with tumor size and position, our detection is. In recent techniques, such as machine learning or deep learning, lung cancer can be predicted but this technique is not appropriate for predicting image segmentation in that particular field.