Probabilistic Neural Network to Classify Image of Children’s Face with Down Syndrome
Romi Fadillah Rahmat, Safrida Budiarti, Sharfina Faza, Insidini Fawwaz, Ulfi Andayani · Journal of Physics Conference Series · 2020
Down Syndrome is a condition of the physical and mental underdevelopment of a child due to chromosomal development abnormalities. This condition has different facial symptoms. Although Down syndrome has a unique feature on the face, they have a similar face with their parents and siblings, so it is quite complex to tell the difference. These symptoms contain specific information for facial recognition. In this research, the proposed method consists of five stages. The first stage is the input image. The second stage is pre-processing which composed of grayscale and CLAHE processes. The third stage is the image segmentation to bring up the special features of the image. The process continued with feature extraction that will generate an invariant moment. The final stage is the classification process to determine the type of face such as mosaic, trisomy 21 or without syndrome. The test result using the test data of 22 images showed that the system has an accuracy rate of 91% with the success percentage for recall and precision is at 88%.