Tongue Image Segmentation using Hybrid Multilevel Otsu Thresholding and Harmony Search Algorithm

Abstract

This paper proposes the use of hybrid multilevel thresholding and harmony search algorithm (HSA) methods for segmenting the image of the tongue. The Otsu thresholding algorithm aided by the HSA optimization algorithm is used to look for optimal solution in order to produce the best threshold based on the inserted class. Parameters used there are three levels, namely minimum, medium, and maximum. The dataset used consists of the image of the Hongkong University Polytechnic Biometric research center and the acquisition image. The result shows that the best HSA parameters value is at the medium level with threshold value 5 for benchmark image with image quality of 35.50 dB, maximum level with threshold value 5 in biometric image with image quality of 36.51 dB, and maximum level with threshold value 5 on image acquisition with image quality of 38.02 dB. Furthermore, it is expected to conduct research with a threshold value of more than 5 with maximum HSA parameters so that it can be used better again in the diagnosis of the disease through the tongue.