Automatic Detection of Fetal Head using Haar Cascade and Fit Ellipse

USG examination is a routine activity during pregnancy with USG machine. There are several parameters to get information about the fetal. The important parameter is head circumference to get to know abnormality, early detection, and monitoring growth. Accurate result and rapidity are really important during the examination. Low quality of ultrasound image and time-consuming in manual annotation made many variability values depending on the doctor or sonographer, the highest risk of human error, so the result is not accurate enough. From that problem, we need an automatic way to detection of the fetal head to make it faster than the manual way. In this result, we use pattern recognition to give learn the system to recognize the characteristic of the fetal head object. We use Haar Cascade to train the classifier to detect the object in the USG image. We also use Fit Ellipse to get the curve shape of the fetal head. The accuracy rate of HC measurement is 97,23% and just consume 1.9s for fetal head detection.