Lung cancer is one of the most precarious dysfunctions to humankind species and amongst the leading causes of human life expiration, especially in developing countries. Mycobacterium Tuberculosis bacterium is a causative agent of lung cancer. The highly aerobic physiology of M. tuberculosis requires suitable oxygen for survival, which is why Lung is its habitat. Lung cancer is fatal because its detection is challenging, especially in smokers. This study presents a machine vision-based approach for lung cancer detection through CT (computerized tomography) scan images. The study aims to ensure reliable, precise, and accurate detection of lung cancer through texture features extracted from CT scan images (acquired from Bahawal Victoria hospital Bahawalpur, Pakistan). Pre-processing techniques (grayscale conversion, filtration, etc.) played an influential role in removing noise, which might reduce accuracy. Mazda tool has been used for feature extraction and identification of 30 optimized features using three techniques F (Fisher) + PA (probability of error + average correlation) +MI (mutual information). The data mining tool Weka has deployed different classification algorithms with ten cross-validation folds. Artificial Neural Network (ANN: n class) showed comparatively better and probably best accuracy of 95.66 %, respectively.