Vision-Based Satellite Recognition and Pose Estimation Using Gaussian Process Regression

Table 7

Joint satellite recognition and pose estimation results on the 1D subset of BUAA-SID 1.5. Results are in bold if the proposed method performs the best and in italic if the second best.

Representation	Method	MAE (°)
BI	Homeomorphic manifold [12]	4.33
	SVM [12]	9.98
	Kernel method [19]	6.80
	Proposed method	4.83
DT	Homeomorphic manifold [12]	7.95
	SVM [12]	13.5
	Kernel method [19]	10.47
	Proposed method	6.78
HU	Homeomorphic manifold [12]	64.0
	SVM [12]	74.7
	Kernel method [19]	84.65
	Proposed method	57.46
FD	Homeomorphic manifold [12]	27.7
	SVM [12]	25.5
	Kernel method [19]	58.34
	Proposed method	21.85
GI	Homeomorphic manifold [12]	0.51
	SVM [12]	3.05
	Kernel method [19]	0.43
	Proposed method	0.29
HOG	Homeomorphic manifold [12]	8.73
	SVM [12]	10.2
	Kernel method [19]	6.47
	Proposed method	5.32