Efficient Feature Extraction and Segmentation Methods Used in Tuberculosis Detection

According to X-ray images, the segmentation and classification with noise removal are the major stages. When seen from different perspectives or in various lighting conditions, the chest image appears differently. This paper discusses the major threat that tuberculosis poses across the globe. Although there are many different medical treatments available, TB diagnosis is still challenging. Several clinical and diagnostic processes in earlier poster anterior chest radiographs contain computationally developed algorithms that simplify scientific diagnostic analysis by utilizing image acquisition. It is possible that a digital image will be required for the analysis and annotation of the patient's demographic information while TB is being screened via chest radiography. Special screening processes for victimization. This work proposed a novel detection model of tuberculosis. Initially, analyzed the input noises after that median filter performing pre-processing followed by watershed mechanism performing segmentation. Next, the feature extraction is carried out via GLCM feature extraction model and the support vector machine (SVM) classifying the normal and tuberculosis. The Kaggle dataset with the MATLAB software handled the implementation part and it describes that the proposed work performances are higher than existing works.