Databases

UGR Hyperspectral Image Database

The UGR Hyperspectral Image Database currently contains fourteen hyperspectral reflectance factor images. They were acquired using a volumetric Bragg-grating based hyperspectral imager: hyperspectral camera V-EOS by Photon etc . in outdoor environment in Granada, Spain, from May 2014 to September 2014. The scene contents include trees, buildings, car, motorbike, etc. Most of the images have spatial resolution of 1000 × 900 pixels. The spectral range is from 400 nm to 1000 nm in 10 nm intervals, resulting in total 61 channels.

The 61-channel hyperspectral images are saved losslesslyao_2015_D15 in tiff image format, 16 bits per channel, and has size about 155M per image. To get the hyperspectral data in the range of [0,1], we need to normalize it by 216 after reading it from the tiff file. For example in Matlab, we can use the following command to read out a 1000 × 900 × 61 data matrix in range [0,1] with single precision: im = imread('scene1_sp.tiff')/216;

RGB images are rendered under CIE Standard Illuminant D65 with the CIE 1931 2° Standard Observer with gamma correction (γ = 0.6). They are saved losslessly in tiff image format, 8 bits per channel, and has size about 2.7M per image.

Scene1 (1000×900)Scene1 Download RGB image Download hyperspectral image Scene2 (1000×900)Scene2 Download RGB image Download hyperspectral image
Scene3 (1000×900)Scene3 Download RGB image Download hyperspectral image Scene4 (1000×900)Scene4 Download RGB image Download hyperspectral image
Scene5 (1000×900)Scene5 Download RGB image Download hyperspectral image Scene6 (1000×900)Scene6 Download RGB image Download hyperspectral image
Scene7 (1000×900)Scene7 Download RGB image Download hyperspectral image Scene8 (1000×900)Scene8 Download RGB image Download hyperspectral image
Scene9 (1000×900)Scene9 Download RGB image Download hyperspectral image Scene10 (1000×900)Scene10 Download RGB image Download hyperspectral image
Scene11 (1000×900)Scene11 Download RGB image Download hyperspectral image Scene12 (900×1000)Scene12 Download RGB image Download hyperspectral image
Scene13 (900×1000)Scene13 Download RGB image Download hyperspectral image Scene14 (900×1000)Scene14 Download RGB image Download hyperspectral image

IMPORTANT: If you use this data, you must cite the source publication in full: "Outdoor scene reflectance measurements using a Bragg-grating-based hyperspectral imager", J. Eckhard, T. Eckhard, E.M. Valero, J.L. Nieves, E. Garrote Contreras, Applied Optics, vol. 54, N. 13, pp. D15-D24 (2015)

Granada daylight spectral database

daylight_database The spectral daylight data available here (2600 daylight spectra) was measured for all sky states during a two year period at Granada, Spain. In the paper J. Hernández-Andrés, J. Romero, J.L. Nieves & R.L. Lee, Jr., "Color and spectral analysis of daylight in southern Europe", Journal of the Optical Society of America A, Vol. 18, N. 6, pp. 1325-1335, June 2001 we describe in detail the chromaticity coordinates, correlated color temperatures (CCT), luminous efficacies, and relative UV and IR contents of Granada daylight. Our principal-component analysis shows that Granada daylight spectra can be adequately represented by using six-dimensional linear models in the visible, whereas seven-dimensional models are required if we include the UV or near-IR. Yet on average only three-dimensional models are needed to reconstruct spectra that are colorimetrically indistinguishable from the original spectra.

IMPORTANT: If you use this data, you must cite the source publication in full: J. Hernández-Andrés, J. Romero, J.L. Nieves & R.L. Lee, Jr., "Color and spectral analysis of daylight in southern Europe", Journal of the Optical Society of America A, Vol. 18, N. 6, pp. 1325-1335, June 2001

Spectral Power Distributions of 2600 daylight spectra (global spectral irradiances on a horizontal surface) measured for all sky states during a 2-year period at Granada, Spain. The spectral range was from 300 nm to 1100 nm daylight.
Spectral range 380-780nm 400-700nm 300-1100nm
Spectral resolution 5 nm (81 rows) 5 nm (61 rows) 5 nm (161 rows)
Original data: Spectral Power Distribution of global irradiance Download MAT fileimage Download MAT fileimage
Non-normalized first six eigenvectors (The eigenvectors were calculated without normalization of the daylight spectra) Download ASCII file image Download ASCII fileimage
Normalized first six eigenvectors (The eigenvectors were calculated from daylight spectra previously normalized using the L2 norm) Download ASCII file image Download ASCII fileimage
Granada standard DAYLIGHT spectra at different correlated colour temperatures. Reconstructed standard daylight spectra from the Granada eigenvectors for different correlated color temperatures.In ASCII data: one spectra per column.
Spectral range 380-780nm
Spectral resolution 5 nm (81 rows)
Correlated color temperatures (Kelvin) 4500, 5500, 6500, 7500, 8500, 9500, 10000 and 10500 K (see first row of the file)
Download ASCII file image

Granada SKYLIGHT spectral database

The spectral skylight data available here (over 1500 skylight spectra) was measured measured during a seven-month period in Granada, Spain. In the paper J. Hernández-Andrés, J. Romero& R.L. Lee, Jr., "Colorimetric and spectroradiometric characteristics of narrow-field-of-view clear skylight in Granada, Spain", Journal of the Optical Society of America A, Vol. 18, N. 2, pp. 412-420, Feb 2001 we use spectral radiances measured within 3° fields of view (FOV's) to define colorimetric characteristics along four sky meridians: the solar meridian and three meridians at azimuths of 45°, 90°, and 315° relative to it. The resulting clear-sky chromaticities in 44 different view directions (1) are close to but do not coincide with the CIE daylight locus, (2) form V-shaped meridional chromaticity curves along it (as expected from theory), and (3) have correlated color temperatures (CCT's) ranging from 3800 K to infinity K. We also routinely observe that sky color and luminance are asymmetric about the solar meridian, usually perceptibly so. A principal-components analysis shows that three vectors are required for accurate clear-sky colorimetry, whereas six are needed for spectral analyses

IMPORTANT: If you use this data, you must cite the source publication in full: J. Hernández-Andrés, J. Romero& R.L. Lee, Jr., "Colorimetric and spectroradiometric characteristics of narrow-field-of-view clear skylight in Granada, Spain", Journal of the Optical Society of America A, Vol. 18, N. 2, pp. 412-420, Feb 2001

image Spectral Power Distributions of 1567 daylight spectra (spectral radiances measured within 3° fields of view (FOV's) at different points of the sky dome at Granada, Spain. The spectral range was from 370 nm to 790 nm and the spectral resolution of 5 nm (85 rows). Download MAT file.

Labial Teeth and Gingiva Image Database - LTG-IDB

The LTG-IDB currently contains 90 photographic digital still images from 27 subjects. Images are available in several sets that differ by resolution, file format and image scene type. This leads to a total of 405 images occupying about 7GB of memory. labial Images in this database are available in raw image format (which is the unprocessed sensor data of the camera in a specific vendor dependent data format, in this case the Canon raw image format cr2), as well as JPEG and TIFF. The strengths of this database are the fixed, well-defined and well known parameters of image acquisition. In addition, the variety of analytical and methodological applications shall be highlighted, as this database includes raw images, as well as pre-processed JPEG and TIFF images and is published under a creative common license.

A detailed description on the properties of this databse is given in: T. Eckhard, E. Valero, J. Nieves, F. Mesa, “The Labial Teeth and Gingiva Photographic Image Database LTG-IDB, technical report”, 2011 .

IMPORTANT: If you use this data, you must cite the source publication in full: Labial teeth and gingiva color image segmentation for gingival health-state assessment, T. Eckhard, E.M. Valero, J.L. Nieves, IS&T/ 6th European Conference on Color in Graphics, Imaging and Vision, CGIV 2012, pp. 102-107, Amsterdam (Netherlands); (2012)

Daylight spectra generated with SBDART

The global radiation spectrum is affected by a wide range of atmospheric components, including solar geometry, aerosols, gases, cloud cover and surface reflectance properties which all modify the spectral radiation as it passes through the atmosphere and reaches ground level. With an aim of understanding the influence of atmospheric components and phases of daylight on colorimetric specifications of downward radiation, we incorporate a wide range of plausible atmospheric parameters into a radiative transfer code and generate a large set of illuminations which may commonly occur in nature. We investigate colorimetric extension of plausible daylight, sunlight and skylight illuminations and analyze atmospheric components which cause substantial variation in the color of natural illumination. josa_2016_1049

You can download the data at: https://drive.google.com/folderview?id=0B7K29Xuy4jdmZ0tIQWI4RDNMSk0

IMPORTANT: If you use this data, you must cite the source publication in full: Colorimetric analysis of outdoor illumination across varieties of atmospheric conditions, S. Peyvandi, J. Hernández-Andrés, F.J. Olmo, J.L. Nieves, J. Romero, Journal of the Optical Society of America A, vol. 33, n. 6, pp. 1049-1059 (2016)