One of the deadliest illnesses that may affect a person is lung cancer. The path of medical treatment for cancer will primarily depend on its nature and location. Early detection of cancer cells has the potential to save a great deal of precious lives. An automated method has to be created in order to detect malignant situations as soon as feasible. Prediction accuracy has always been a problem, despite the different approaches that have been put out in the past by numerous academics. In this article, a convolutional neural network (CNN) based method for identifying abnormal lung tissue development is proposed. To achieve high accuracy, a tool that has a higher chance of being detected is taken into account. Histological images of biopsied tissue from potentially infected lungs are used by medical professionals to diagnose patients. It takes a long time and is often prone to mistakes to diagnose different types of lung cancer. CNN has improved speed and accuracy in identifying and classifying different types of lung cancer, which helps to improve patient outcomes by identifying the most effective course of therapy. We utilized the “Histopathological images” Kaggle dataset. Lung adenocarcinomas, lung squamous cell carcinomas, and lung malignancies of the normal type are all included in this study. The primary components of this technique are model selection, assessment, pre-processing, and data collecting. The accuracy rates for the training and validation of the CNN model were 97.5 and 98.5%, respectively.