BibTex Citation Data :
@article{YPJ2592, author = {Nelly Mirnasari and Kusworo Adi}, title = {APLIKASI METODE OTSU UNTUK IDENTIFIKASI BAKTERI TUBERKULOSIS SECARA OTOMATIS}, journal = {Youngster Physics Journal}, volume = {2}, number = {2}, year = {2013}, keywords = {}, abstract = { Tuberculosis (TB) is an infectious disease caused by Mycobacterium tuberculosis. In this research, Ziehl-Neelsen sputum image is used to identify TB bacteria. Image segmentation process is done by using Otsu auto-thresholding to split object from its background in the ZN-stain sputum sample image. Morphology operation is done toward the splitting result to obtain the value of compactness, metric, and eccentricity of each object in the image which later will be used as input for the Adaptive Neuro-Fuzzy Inference System (ANFIS) to be classified. Result of the research show that segmentation using Otsu method is capable to properly split the object expected as bacteria from its background. The network testing show that ANFIS architecture with gbellmf as membership function type will work properly if it has 2 MFs on each input with epoch of 40. Keywords: Tuberculosis, Ziehl-Neelsen (ZN) Stain, Otsu method, auto-thresholding, fuzzy, ANFIS. }, issn = {2302-7371}, pages = {13--20} url = {https://ejournal3.undip.ac.id/index.php/bfd/article/view/2592} }
Refworks Citation Data :
Tuberculosis (TB) is an infectious disease caused by Mycobacterium tuberculosis. In this research, Ziehl-Neelsen sputum image is used to identify TB bacteria. Image segmentation process is done by using Otsu auto-thresholding to split object from its background in the ZN-stain sputum sample image. Morphology operation is done toward the splitting result to obtain the value of compactness, metric, and eccentricity of each object in the image which later will be used as input for the Adaptive Neuro-Fuzzy Inference System (ANFIS) to be classified.
Result of the research show that segmentation using Otsu method is capable to properly split the object expected as bacteria from its background. The network testing show that ANFIS architecture with gbellmf as membership function type will work properly if it has 2 MFs on each input with epoch of 40.
Keywords: Tuberculosis, Ziehl-Neelsen (ZN) Stain, Otsu method, auto-thresholding, fuzzy, ANFIS.
Last update:
