Human Head Angle Detection Based on Image Analysis

The paper presents head angle recognition problem which has been widely studied in computer vision and takes an important role in different applications. We consider the driver monitoring application that uses the head angle detection for distraction dangerous state detection. Base method that can be used is embedded to Dlib but unfortunately, it provides possibilities to determine small values of angles. If the face is turned more than 35° Dlib detector lose it. We proposed and compared two methods to solve this problem. We divide the head angle recognition problem into two parts. The first one is driver face detection and the second one is to find points in the face to estimate rotation angle. We propose the method that allows to determine angle of the rotated driver head up to 70°. The method was successfully implemented to the driver monitoring system and tested.