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

Table 1

The self-occlusion avoidance process of visual object Bunny with translation.

Information	Image number
	1	2	3	⋯	6	⋯	10	⋯	16	17
Depth image				⋯		⋯		⋯
Matching Gaussian curvature feature matrices with SIFT algorithm	—			⋯		⋯		⋯		—
Total matched points	—	221	149	⋯	214	⋯	305	⋯	323	—
Error matched points	—	32	33	⋯	47	⋯	55	⋯	35	—
Error rate	—	14.5%	22.2%	⋯	21.9%	⋯	18.0%	⋯	10.8%	—
Observation position	0.00, −1.00, 300.00	0.00, −1.00, 300.00	60.05, −3.71, 293.06	⋯	154.94, −60.20, 243.36	⋯	222.52, −104.03, 167.34	⋯	256.37, −122.16, 94.13	260.57, −122.69, 88.72
Observation direction	0.00, 1.00, −300.00	0.00, 1.00, −300.00	−54.56, −1.73, −294.99	⋯	−168.23 95.91, −229.13	⋯	−245.97, 120.77, −122.12	⋯	−258.67, 123.17, −89.00	−265.93, 123.99, −62.51
Observed area	—	167.61	587.02	⋯	2158.60	⋯	2625.68	⋯	3014.34	3025.58