Tuberculosis (tb) Identification in the Ziehl-neelsen Sputum Sample in Ntsc Channel and Support Vector Machine (svm) Classification

Tuberculosis (TB) is an important public health issue in this world. In 1992, World Health Organization (WHO) has determined tuberculosis as Global Emergency. The cause of tuberculosis disease is Mycobacterium tuberculosis. This bacteria is bar shaped and has the quality of acid fast so that it’s also known as Acid Fast Bacilli (AFB). In this research, an algorithm to identifying and counting the number of tuberculosis bacteria has developed by microscope imaging. Color segmentation is done by way of extracting the Saturation channel of NTSC (Luminance, Hue, saturation) color model. Thresholding process is using Otsu Method. Feature extraction for bacteria shape identification process using two parameters i.e. eccentricity and compactness. The training and object recognition using Support Vector Machine algorithm. The result of development counting is equal to the manual counting result. This is show that Support Vector Machine is good to be applied in detecting and counting the number of tuberculosis bacteria.