Description

ICVL is a hyperspectral image dataset, collected by "Sparse Recovery of Hyperspectral Signal from Natural RGB Images"

The database images were acquired using a Specim PS Kappa DX4 hyperspectral camera and a rotary stage for spatial scanning. At this time it contains 200 images and will continue to grow progressively.

Images were collected at 1392 $\times$ 1300 spatial resolution over 519 spectral bands (400-1,000nm at roughly 1.25nm increments). The .raw files contain raw out-of-camera data in ENVI format and .hdr files contain the headers required to decode them. For your convenience, .mat files are provided, downsampled to 31 spectral channels from 400nm to 700nm at 10nm increments.

The original dataset only contains clean images. For hyperspectral image denoising benchmarks, the testing datasets come from "3D Quasi-Recurrent Neural Network for Hyperspectral Image Denoising".

Quick look



4cam_0411-1640-1	4cam_0411-1648	bguCAMP_0514-1659	bguCAMP_0514-1711

bguCAMP_0514-1712	bguCAMP_0514-1718	bguCAMP_0514-1723	bguCAMP_0514-1724

BGU_0403-1419-1	bgu_0403-1439	bgu_0403-1444	bgu_0403-1459

bgu_0403-1511	bgu_0403-1523	bgu_0403-1525	BGU_0522-1113-1

BGU_0522-1127	BGU_0522-1136	BGU_0522-1201	BGU_0522-1203

BGU_0522-1211	BGU_0522-1216	BGU_0522-1217	bulb_0822-0903

bulb_0822-0909	CC_40D_2_1103-0917	eve_0331-1549	eve_0331-1551

eve_0331-1601	eve_0331-1602	eve_0331-1606	eve_0331-1618

eve_0331-1632	eve_0331-1633	eve_0331-1646	eve_0331-1647

eve_0331-1656	eve_0331-1657	eve_0331-1702	eve_0331-1705

Flower_0325-1336	gavyam_0823-0930	gavyam_0823-0933	gavyam_0823-0944

gavyam_0823-0945	gavyam_0823-0950-1	grf_0328-0949	hill_0325-1219

hill_0325-1228	hill_0325-1235	hill_0325-1242	IDS_COLORCHECK_1020-1215-1

IDS_COLORCHECK_1020-1223	Labtest_0910-1502	Labtest_0910-1504	Labtest_0910-1506

Labtest_0910-1509	Labtest_0910-1510	Labtest_0910-1511	Labtest_0910-1513

lehavim_0910-1600	lehavim_0910-1602	lehavim_0910-1605	lehavim_0910-1607

lehavim_0910-1610	Lehavim_0910-1622	Lehavim_0910-1626	Lehavim_0910-1627

Lehavim_0910-1629	Lehavim_0910-1630	Lehavim_0910-1633	Lehavim_0910-1635

Lehavim_0910-1636	Lehavim_0910-1640	Lehavim_0910-1708	Lehavim_0910-1716

Lehavim_0910-1717	Lehavim_0910-1718	Lehavim_0910-1725	lst_0408-0950

lst_0408-1004	lst_0408-1012	Master20150112_f2_colorchecker	Master2900k

Master5000K	Master5000K_2900K	Maz0326-1038	maz_0326-1048

mor_0328-1209-2	nachal_0823-1038	nachal_0823-1040	nachal_0823-1047

nachal_0823-1110	nachal_0823-1117	nachal_0823-1118	nachal_0823-1121

nachal_0823-1127	nachal_0823-1132	nachal_0823-1144	nachal_0823-1145

nachal_0823-1147	nachal_0823-1149	nachal_0823-1152	nachal_0823-1210-4

nachal_0823-1213	nachal_0823-1214	nachal_0823-1217	nachal_0823-1220

nachal_0823-1222	nachal_0823-1223	negev_0823-1003	negev_0823-1005

objects_0924-1550	objects_0924-1556	objects_0924-1557	objects_0924-1558

objects_0924-1600	objects_0924-1601	objects_0924-1602	objects_0924-1605

objects_0924-1607	objects_0924-1610	objects_0924-1611	objects_0924-1612

objects_0924-1614	objects_0924-1617	objects_0924-1619	objects_0924-1620

objects_0924-1622	objects_0924-1628	objects_0924-1629	objects_0924-1631

objects_0924-1632	objects_0924-1633	objects_0924-1634	objects_0924-1636

objects_0924-1637	objects_0924-1638	objects_0924-1639	objects_0924-1641

objects_0924-1645	objects_0924-1648	objects_0924-1650	objects_0924-1652

omer_0331-1055	omer_0331-1102	omer_0331-1104	omer_0331-1118

omer_0331-1119	omer_0331-1130	omer_0331-1131	omer_0331-1135

omer_0331-1150	omer_0331-1159	peppers_0503-1308	peppers_0503-1311

peppers_0503-1315	peppers_0503-1330	peppers_0503-1332	pepper_0503-1228

pepper_0503-1229	pepper_0503-1236	plt_0411-1037	plt_0411-1046

plt_0411-1116	plt_0411-1155	plt_0411-1200-1	plt_0411-1207

plt_0411-1210	plt_0411-1211	plt_0411-1232-1	prk_0328-0945

prk_0328-1025	prk_0328-1031	prk_0328-1034	prk_0328-1037

prk_0328-1045	rsh2_0406-1505	rsh_0406-1343	rsh_0406-1356

rsh_0406-1413	rsh_0406-1427	rsh_0406-1441-1	rsh_0406-1443

sami_0331-1019	sat_0406-1107	sat_0406-1129	sat_0406-1130

sat_0406-1157-1	selfie_0822-0906	strt_0331-1027	tree_0822-0853

ulm_0328-1118

Credits

Dataset originally collected by ICVL from the webpage: https://icvl.cs.bgu.ac.il/hyperspectral/

For questions, comments and technical assistance, please contact [email protected]

When used, fully or partially, please cite:

Arad and Ben-Shahar, Sparse Recovery of Hyperspectral Signal from Natural RGB Images, in the European Conference on Computer Vision, Amsterdam, The Netherlands, October 11–14, 2016

Bibtex:

@inproceedings{arad_and_ben_shahar_2016_ECCV,
  title={Sparse Recovery of Hyperspectral Signal from Natural RGB Images},
  author={Arad, Boaz and Ben-Shahar, Ohad},
  booktitle={European Conference on Computer Vision},
  pages={19--34},
  year={2016},
  organization={Springer}
}