Deteksi Fitur Wajah pada Hasil Usg 3D Janin dengan Algoritma Viola-Jones

Abstract

Ultrasound (Ultrasonography) is a medical imaging technology used for monitoring fetal development during pregnancy. It produces visual images that support the process of medical diagnosis and pregnancy monitoring. Image processing is used to process and analyze digital images, providing valuable additional information. One of the methods used in image processing is the Viola-Jones algorithm. The Viola-Jones algorithm is commonly used to detect objects in image processing and computer vision, especially human faces. Face detection using the Viola-Jones algorithm can be implemented using software such as Matlab and Python. However, in this experiment, there were difficulties in using the Viola-Jones algorithm to detect baby ultrasound images. This is caused by the position seen from the side in the baby's ultrasound image and the uniformity of color throughout the face. In fact, some features such as eyebrows and eyes are difficult to recognize clearly. In addition, the resulting training files may not provide consistent results on other samples due to complex variations in infant ultrasound images.