Mycobacterium Tuberculosis is the causative agent of the transferable disease called tuberculosis (TB). Early diagnosis of TB via sputum examination is imperative in an effort to avoid transmission. Microscopic examination of sputum involves observing 100-300 fields of view (FoV) by eye, which typically takes 30-150 minutes. In practice, pathologists still manually change the FoV to cover the entire sample. This process is associated with limitations like low accuracy in identifying local features due to reduced contrast in the images, misregistration in overlapping regions, and large computation times. To correct these limitations, a digitization system in the form of whole slide imaging (WSI) is required, this method drastically minimizes examination time, with WSI bacterial detection requiring only 3-10 minutes, which is a significant improvement in diagnostic effectiveness without compromising the completeness of the analysis. In the proposed framework, invariant local features are identified using the Scale-Invariant Feature Transform (SIFT) algorithm, K-Nearest Neighbors (K-NN) and Brute Force (BF) Matcher are used in the proposed algorithm to match features precisely. Its ability to produce permanently aligned composite images is further reflected in the mosaiced result's zero-pixel measurement, which achieved at least 1,069,687 pixels.