Best of both worlds: Fusing hyperspectral data from two generations of spectro-imagers for X-ray astrophysics

Home

All issues

Volume 694 (February 2025)

A&A, 694 (2025) A34

Full HTML

Open Access

Table 2

Error metric results comparing regularizations for the four toy models.

Gaussian (0.5–1.4 keV)

Regularization	NMSE	aSAM	acSSIM	ERGAS

W2D1D	21.36	0.571	0.017	0.220
LRS	20.97	0.557	0.011	0.116
LRW2D	21.01	0.560	0.012	0.114

Gaussian (6.2–6.9 keV)

	NMSE	aSAM	acSSIM	ERGAS

W2D1D	18.67	0.599	0.005	0.229
LRS	14.16	0.600	0.011	0.343
LRW2D	16.57	0.609	0.010	0.248

Gaussian with Rebinning (0.5–1.4 keV)

	NMSE	aSAM	acSSIM	ERGAS

W2D1D	10.36	0.691	0.048	0.422
LRS	11.01	0.670	0.045	0.341
LRW2D	10.34	0.690	0.054	0.317

Realistic with rebinning (0.5–1.4 keV)

	NMSE	aSAM	acSSIM	ERGAS

W2D1D	9.956	0.700	0.032	0.608
LRS	10.16	0.691	0.031	0.358
LRW2D	9.523	0.712	0.038	0.351

Notes. The NMSE (Equation (A.1)) is better if high, while the aSAM (Equation (A.2)) and acSSIM (Equation (A.3)) are better if low. The best values are marked in bold.

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.