Table 1 Properties of five open-source hyperspectral datasets.
From: A survey on computational spectral reconstruction methods from RGB to hyperspectral imaging
Dataset | Amount | Resolution | Spectral channels | Spectrum/(nm) | Featured scenes |
|---|---|---|---|---|---|
CAVE34 | 32 | \(512 \times 512\) | 31 | 400–700 | Skin, hair, food and drink |
ICVL29 | 203 | \(1392 \times 1300\) | 31 | 400–700 | Urban, rural, indoor and plant |
BGU-HS27 | 286 | \(1392 \times 1300\) | 31 | 400–700 | Urban, rural, indoor and plant |
ARAD-HS28 | 510 | \(512 \times 482\) | 31 | 400–700 | Statue, vehicle and paint |
KAUST-HS26 | 409 | \(512 \times 512\) | 34 | 400–730 | Vehicle, food, building and toy |
