Head attitude tracking device and method based on incomplete characteristics

Abstract

The invention discloses a head attitude tracking device and method based on incomplete characteristics. Normalized processing on a video stream acquired in real time is carried out by a head attitudedetection system based on computer vision, then whether a face portion exists is determined through characteristic confirmation by a model after CNN training, if face detection fails, and the incomplete facial characteristic part is completed by a trained confrontation network model; the position of the face is tracked in the video, the facial characteristic point position is marked by a sparsely-constrained model, the PnP algorithm is utilized to acquire Pitch, Yaw, Roll three angle values [thetax thetay thetaz] of the head attitude, and the angle values are then sent to a follow-up holder. The method is advantaged in that the accurate head position and attitude information can be provided, and high timeliness and good robustness are realized.

Description

technical field [0001] The invention belongs to the technical field of target detection, tracking and remote control, and more specifically relates to a head posture tracking device and method based on incomplete features. Background technique [0002] Head posture detection is widely used in virtual reality and human-computer interaction, intelligent robot control and fatigue driving detection and other fields. [0003] Statistics show that 25%-30% of road traffic accidents are caused by driver distraction. Studying the driver's head posture, and then analyzing the driver's attention, will help reduce the incidence of traffic accidents. Therefore, head pose detection is of great significance for analyzing the driver's driving behavior and the driver's mental state. [0004] Virtual reality application is a human-computer interaction technology that highly realistically simulates human-computer interaction behaviors such as seeing, hearing, and moving in a natural environm...

AI Technical Summary

Problems solved by technology

[0007] 1. Wearing a motion sensor method: This method installs an accelerometer and a gyroscope sensor on the head of the test subject, and judges the head posture by receiving the sensor motion data. This method requires wearing a device, the user experience is poor, and there is zero-degree drift The problem is that it needs to be calibrated before each use, and the operation is cumbersome

[0008] 2. LED calibration method: By wearing a headgear with a certain number of LEDs, use the camera to capture the headgear image, and then judge the current position according to the position of the LEDs. Similarly, this method has high precision, but still needs to wear the device and perform calibration

Images