Fig. 1. Imaging principle of stereo camera

Fig. 2. Visible light circular array target. (a) Halcon calibration board; (b) circular array calibration board

Fig. 3. Hardware physical diagram of near-infrared target. (a) 850 nm LED beads; (b) printed circuit board

Fig. 4. Near-infrared target imaging effect and comparison of connected component area. (a) Grayscale image; (b) three-dimensional color mapping; (c) comparison of connected area of the LEDs

Fig. 5. Preprocessing effect of feature points. (a) Original image of large scale; (b) preprocessing result of large scale; (c) original image of small scale; (d) preprocessing result of small scale

Fig. 6. Ideal edge model. (a) Original edge model; (b) rotated edge model

Fig. 7. Ellipse in image coordinate system

Fig. 8. Algorithm process of outlier edges removal

Fig. 9. Near-infrared target in world coordinate system

Fig. 10. Schematic of feature points sorting

Fig. 11. Sorting results at different rotation angles

Fig. 12. Simulated near-infrared target projection image. (a) Target projection image; (b) local small scale; (c) local large scale

Fig. 13. Pixel error under different standard deviations of Gaussian white noise

Fig. 14. Near-infrared stereo camera system and experimental schematic diagram. (a) System in photographing target; (b) schematic of photographing target

Fig. 15. Distributions of stereo mean reprojection error for five algorithms

Fig. 16. Distribution of mean reprojection error

Fig. 17. Impact of three parameters on mean reprojection error. (a) kt; (b) lt; (c) σ

Fig. 18. Target reconstruction result

Fig. 19. Statistical histogram of samples. (a) Before removing outliers; (b) after removing outliers