Dynamic Self-Occlusion Avoidance Approach Based on the Depth Image Sequence of Moving Visual Object

Table 3

The self-occlusion avoidance process of visual object Duck with both translation and rotation.

Information	Image number
	1	2	3	⋯	6	⋯	9	⋯	12	13
Depth image				⋯		⋯		⋯
Matching Gaussian curvature feature matrices with SIFT algorithm	—			⋯		⋯		⋯		—
Total matched points	—	552	431	⋯	456	⋯	432	⋯	438	—
Error matched points	—	20	20	⋯	38	⋯	22	⋯	71	—
Error rate	—	3.6%	4.6%	⋯	8.3%	⋯	5.1%	⋯	16.2%	—
Observation position	0.00, −1.00, 300.00	0.00, −1.00, 300.00	−0.18, 26.13, 297.59	⋯	−9.32, 65.62, 292.07	⋯	−19.88, 106.03, 279.18	⋯	−49.74, 135.68, 265.74	−24.30, 105.82, 258.45
Observation direction	0.00, 1.00, −300.00	0.00, 1.00, −300.00	0.84, −25.99, −298.87	⋯	14.72, −68.07, −291.81	⋯	17.58, −92.39, −284.88	⋯	23.96, −120.71, −273.60	56.05, −114.93, −271.39
Observed area	—	83.04	222.32	⋯	317.65	⋯	462.84	⋯	705.77	733.58