<![CDATA[ABSTRACT The face is a source of biometric technology that characterizes the body parts attached to a person, which is often the object of research in the field of digital image processing. Facial bio-metric technology has been widely used in various electronic devices for facial recognition. However, the first step in facial biometrics is face detection. The purpose of face detection is to find facial areas in the image. Face detection is done by using a computer process in learning to recognize certain features, such as facial features, eyes, nose position, and lip shape. The computer will present a face image in the form of a two-dimensional matrix or more. The goal is to find out whether or not there is a face on an object in security verification. So, research related to face detection is a fundamental topic and plays an important role in the development of biometric technology. This study applies the Viola-Jones method to detect the presence of a human face with a variety of facial poses and accessories that are often worn on the face. This method has high accuracy and simple feature classification. This method does not use pixels directly in classifying images but instead uses the Haar feature to combine Integral Image and AdaBoost Machine Learning calculations to form a Cascade Classifier so that the results for the presence of faces are obtained. The research implementation uses Android and face detection is taken from screenshots stored on the Android device. The results of testing with an accuracy rate of up to 95.38% with an upright face position facing the camera and images with accessories an accuracy rate of up to 72.47%. This shows that the use of accessories can cover some facial features, thus making it difficult for the face detection algorithm to recognize faces accurately. Keywords: Android, Face Detection, Face Variation, Viola Jones]]>
