Klasifikasi Citra pada Wajah Down Syndrome dengan Support Vector Machine

Abstract

Disorders that affect physical and mental growth retardation are known as Down Syndrome. The characteristics seen in patients with the syndrome are not like normal conditions in general, such as relatively short height, small head size, flat nose shape similar to Mongolian people. Down Syndrome is a multiple congenital disorder caused by excess genetic material on chromosome 21. The uniqueness of Down Syndrome can be seen in the difference in the number of chromosome combinations in the body which has three types, namely Trisomy-21, Translocation and Mosaic. The three types of Down Syndrome have similar physical characteristics when seen in person. However, the disorder can only be seen through the genetics that exist in a person. With this background, a technological approach with the Support Vector Machine (SVM) method is needed in classifying the types of Down Syndrome which will be processed through digital image processing. The stages of digital image processing are resizing, grayscalling and median filter. Continued using the canny edge thresholding method and the zoning method for feature extraction. The image will be classified according to its type using the Support Vector Machine. This study consists of 231 data which were processed into two types of data; training data and testing data. Based on the test result, the system was able to classify the type of Down Syndrome with an accuracy of 82%.