Laser & Optoelectronics Progress, Volume. 57, Issue 18, 181008(2020)
Head Pose Estimation Algorithm Based on Structured Light Three-Dimensional Reconstruction
Fig. 1. Head pose. (a) Flow chart of head pose estimation; (b) world and intrinsic coordinate system
Fig. 2. Estimate the angle α.(a) (b) Detect nose tip O and nose bridge B; (c) calculate the angle α between xy plane and b axis
Fig. 3. Estimate the yaw angle. (a) (b) Search the equidistant points from the nose tip on both side of the nose; (c) horizontal cross flat through the nose tip
Fig. 4. Deep image of side face
Fig. 5. Experiment image. (a) Experimental system; (b) light path
Fig. 6. Flowchart of human face 3D reconstruction
Fig. 7. Measurement error of head pose estimation. (a) Yaw angle mean absolute error; (b) pitch angle mean absolute error; (c) roll angle mean absolute error
Get Citation
Copy Citation Text
Junyu Zhong, Jian Qiu, Peng Han, Kaiqing Luo, Li Peng, Dongmei Liu. Head Pose Estimation Algorithm Based on Structured Light Three-Dimensional Reconstruction[J]. Laser & Optoelectronics Progress, 2020, 57(18): 181008
Category: Image Processing
Received: Jan. 6, 2020
Accepted: Feb. 10, 2020
Published Online: Sep. 2, 2020
The Author Email: Qiu Jian (qiuj@scnu.edu.cn), Han Peng (hanp@scnu.edu.cn)