PENGEMBANGAN SISTEM IDENTIFIKASI JENIS KELAMIN JANIN PADA CITRA USG

Abstract

This study aimed at: (1) designing the identification system of fetus genders’ at USG image, and (2) implementing the identification system of fetus genders’ at USG image. In designing and implementing the system, it was used three main types of operation methods, namely segmentation, features extraction, and data clustering. Input of these applications were image with extent bitmap (*.bmp), and its output is an information about fetus genders’. The test has been done on the entire image sample which has become a knowledge based. In the process of testing, it was initially needed the helping of an obstetrician in identifying the fetus genders’. In designing and implementing the application design, it was used a waterfall method which is usually called as a classic life cycle model. This is a classic model which creates the software systematically and sequentially that includes some stages namely: (1) requirements definition, (2) system and software design, (3) implementation and unit testing, and (4) integration and testing system.  The implementation and testing in this study was an identification system of fetus genders’ at USG image that were using a Delphi program. Based on the data of the performance system test about the accuracy of the system when identified the fetus genders’, it was found that until 66.67% could be identified well from 54 USG image sample. In this case, the identification system of the fetus genders’ was very helpful for obstetricians when they would be identifying the fetus genders’.