Menghitung Jumlah Sel Goblet Usus Ayam Secara Otomatis Dengan Metode Multilevel Thresholding

Abstract

Goblet cell was an indicator the health of chicken intestine that was strongly influence of chicken productivity. So far, the determination of goblet cell was conducted by visual used of microscope. These methods required a long time and also the result was subjective value. The aim of this study was image processed for calculating the number of goblet cells in the chicken intestine and assessed the method applied performance. The material used was 30 images of broiler intestine preparations with AB-PAS staining. The research was carried out used multilevel thresholding and labeling methods for calculating the number of goblet cells in intestine of chicken. Chicken intestinal tissue that had been prepared with AB-PAS staining was acquired used a microscope equipped with a camera and connected to a PC. The result of the acquisition was an RGB scale image, so needed to be converted into an HSV scale image. Next, conducted the extraction process the hue component to convert the image to a gray scale. The gray scale image can be used in the process of identifying goblet cells using the multilevel thresholding method. Objects in the form of goblet cells that have been separated from other tissues were then repaired using morphological operations. The results of morphological operations were then calculated using the labeling method. The results showed that the method applied was successful in automatically calculated the number of goblet cells. The results of calculations using the multilevel thresholding and labeling method had accuracy rate of 90% compared with the results of calculated of visually directly by analyst.