Determination of RGB in Fingernail Image As Early Detection of Diabetes Mellitus

Abstract

The aim of this research was to determine of component color RGB on fingernails as early detection of diabetes mellitus. Methods of the study consisted of material preparation and implementation procedures that carried out in three step i.e (1) data retrieval, (2) data processing and (3) data analysis. Firstly, random blood glucose levels were take with Autocheck GCU rapid test then fingernail images data were taken by digital camera and classified into into three categories namely diabetes, prediabetes and normal. Images data were segmented and transformed manually into R (red), G (green), and B (blue) histogram. RGB histogram was analyzed and grouped by frequency distribution to obtain RGB range number of each category. The results showed that range number of Red in diabetes, prediabetes and normal were 160–181, 170–185, and 165–183. Range number of Green were 100–119, 103–123, 107–129 for diabetes, prediabetes and normal. Also range number of Blue were 93–113, 90–110 and 97–117 for diabetes, prediabetes and normal. As conclusion, there was overlapping range number of RGB in all categories. Therefore, fingernail image as early detection of Diabetes Mellitus need to improve by added some feature such as texture image.