Automatic Detection of Tuberculosis using Image Processing and KNN Classifier

Tuberculosis is the biggest infectious disease in human that mainly affects lungs, this tuberculosis is one the top ten prominent cause of death for human. When any patient is suffer from the tuberculosis firstly the person is not aware about he/she is affected with tuberculosis or not because it is a type of disease whose maximum symptoms are very much normal and related to other disease also. So without testing it is not sure that patient has TB. But, it is curable or not a untreated problem. We can overcome from this problem if we can detect it at their earlier stage and save the number of lives because at higher stage medicine getting more costly and patients require to treat and observe more carefully. In this paper proposed a technique of image segmentation and machine learning to diagnosis and classify the types of the tuberculosis (TB). These tools and techniques are used to prior expose of tuberculosis (TB) disease. Also classify the type of TB. The type and the stage of tuberculosis show accuracy 88% of system.