Explore on Doctor's Head Orientation Tracking for Patient's Body Surface Projection Under Complex Illumination Conditions

Doctors head orientation tracking is particularly important in method of augmented reality patients body surface projection. Position and shape of projected image on patients body surface can be adjusted according to such method. Face detection is the basis of head orientation tracking. In view of complex illumination conditions during operation, there are re-detection or false detection in face detection. Based on combination of Kinect depth image and color image, a method of head orientation tracking was presented. Firstly, during training stage, typical AdaBoost algorithm was combined with depth image and Haar-like features were extracted to construct cascade separators, that was used to remove background interference. During detection stage, depth information was used to filter sub-windows to reduce error of re-detection or false detection. Secondly, MM algorithm was used to extract facial feature points. And then a head orientation model was established based on eye and lip feature points, so actual data and BIM Kinect Head Pose database were used to test detection rate, accuracy and resolution. Experimental results show that face detection rate of this algorithm is 98.2%, and false detection rate is 0 under complex illumination conditions. Accuracy of head orientation tracking is 97.2%, and spatial resolution of tracking process is five times higher than that of existing methods.