Spectral Reflectance Estimation Research Articles

Maximum a posteriori estimation of spectral reflectance from color image and multipoint spectral measurements

Accurate color image reproduction under arbitrary illumination can be realized if the spectral reflectance functions in a scene are obtained. Although multispectral imaging is one of the promising methods to obtain the reflectance of a scene, it is expected to reduce the number of color channels without significant loss of accuracy. This paper presents what we believe to be a new method for estimating spectral reflectance functions from color image and multipoint spectral measurements based on maximum a posteriori (MAP) estimation. Multipoint spectral measurements are utilized as auxiliary information to improve the accuracy of spectral reflectance estimated from image data. Through simulations, it is confirmed that the proposed method improves the estimation accuracy, particularly when a scene includes subjects that belong to various categories.

On the remote monitoring of Karenia brevis blooms of the west Florida shelf

In situ surveys (1997–2002) of Karenia brevis distribution on the west Florida shelf were used to explain spectral remote sensing reflectance, chlorophyll- a concentration, and backscattering coefficient estimates derived using SeaWiFS satellite data. Two existing approaches were tested in an attempt to differentiate K. brevis blooms from other blooms or plumes. A chlorophyll-anomaly method used operationally by the National Oceanic and Atmospheric Administration (NOAA) sometimes correctly identified K. brevis blooms but also generated false positives and false negatives. The method identified approximately 1000 km 2 of high chlorophyll-anomalies (>1 mg m −3) off southwest Florida between the 10 and 50-m isobaths nearly every day from summer to late fall. Whether these patches were K. brevis blooms or not is unknown. A second method used a backscattering:chlorophyll- a ratio to identify K. brevis patches. This method separated K. brevis from other blooms using in situ optical data, but it yielded less satisfactory results with SeaWiFS data. Spectral reflectance ( R rs) estimates for K. brevis blooms, diatom blooms, and coastal river plumes are statistically similar for many cases. Large pixel size, shallow water, and imperfect algorithms distort satellite retrievals of bio-optical parameters in patchy blooms. At present, a combination of chlorophyll- a, chlorophyll-anomaly, backscattering:chlorophyll- a ratio, RGB composites, MODIS fluorescence data, as well as time-series analysis and ancillary data such as winds, currents, and sea surface temperature can improve K. brevis bloom assessments. Progress in atmospheric correction and bio-optical inversion algorithms is required to help improve capabilities to monitor K. brevis blooms from space. Further, satellite sensors with improved radiometric capabilities and temporal/spatial resolutions are also required.

分光推定法に基づく画像投影型カラーシミュレータの精度向上

We have proposed an improved color control method for a projector-based color simulator.We use low dimensional response,which is measured with an RGB colorimetric camera to estimate spectral reflectance.In our previous work,a color simulator with a projection image was controlled color change on the basis of the ink supply.By overlapping the controlled projection image based on the spectral reflectance,this simulator helps operator to adjust the color of print more accurately.However,measuring the spectral reflectance from every point on print is expensive.To decrease the cost of spectral measurement,we propose a way of estimating method for spectral reflectance.Only three bands responses were used to estimate the spectral reflectance for the projector-based color simulator.The projection image calculated by approximate spectral reflectance reproduces the color change of the print instead of ink supply.Our color simulator with a projection image matched color accurately;for 32 samples,the average color difference ΔE*94=1.05 and the maximum color difference ΔE*94=2.58.

Estimation of spectral reflectance of object surfaces with the consideration of perceptual color space

Due to the nonlinear transform from spectral reflectance to CIELAB values, the solution obtained by minimizing reflectance error in traditional spectral characterization methods will not be optimal when evaluated by colorimetric error. We propose a reflectance estimation method with consideration of perceptual color space. It combines two approaches, i.e., colorimetric-based spectral calculation and weighted spectral calculation. The experimental results show that the proposed method performs better than previous methods in terms of color difference with very slight degradation in spectral accuracy.

Thin Film Optical Parameters Determination by the Dynamical Modelling and Stochastic Optimization Method

We report on a new method of experimental data processing to obtain optical parameters of thin films. Dynamical modelling of the spectral reflectance can be performed interactively in a graphical environment by the genetic search in wide interval of parameter space and then refined by the genetic algorithm method, by the Nelder-Mead downhill simplex method or Marquardt-Levenberg method. Optical parameters of hydrogenated amorphous silicon (a-Si:H) thin film used in solar cell technology are determined by this new method. The spectral reflectance is a function of optical properties and the thickness of the film. Optical parameters found by our approach do not depend on the initial spectral reflectance estimation.

Spectral Reflectance Estimation of Human Skin and Its Application to Image Rendering

The analysis and synthesis of human skin color is important in many application areas. The skin color depends on some histological variables such as the concentration of pigments melanin and hemoglobin in skin layers. The present paper proposes a method for estimating the surface spectral reflectance of human skin based on an optics model and applying the estimates to 3D realistic image rendering for a human hand. The human skin is modeled as the two layers of the turbid materials for the epidermis and dermis. An estimation algorithm for the two layer model is then developed using the Kubelka–Munk theory. The parameters representing the concentration of pigments are determined based on spectral reflectance measurements of the human skin surface. In the application step, we describe a technique for rendering realistic skin images of a skin surface as a 3D object. The Torrance–Sparrow model is adapted in the image rendering process. The accuracy of the estimated reflectances is shown in experiments, and skin color images are created under a variety of illumination, viewing, and pigmentation conditions.

Estimation of spectral reflectance of snow from IRS-1D LISS-III sensor over the Himalayan terrain

The sensor onboard the satellite views the earth as a plain surface and consequently the satelliteobtained spectral radiances cannot represent true values over a mountainous terrain. The relative magnitudes of terrain slope and its aspect with respect to the sun's position will determine the amount of direct solar radiation incident on an undulating surface. Estimation of spectral reflectance from satellite data forms an important component in many of the snow and glacier studies. The spectral reflectance of snow is influenced by its various parameters. The changes in snowpack characteristics as a result of various metamorphic processes, with age, can cause variations in its spectral reflectances. Since, the terrain geometry also modifies the amount of reflected radiation from a rugged surface, one has to correct the estimated spectral reflectances for terrain topography so as to use them in deriving the snowpack characteristics accurately. Also, the amounts of melt runoff originating from glaciers having different slopes and orientations will not be the same. Considering these aspects, a model has been developed to estimate the terrain corrected spectral reflectances over the Himalayan terrain using the Linear Imaging Self Scanner-III data of the Indian Remote Sensing Satellite. The model computes spectral reflectances from satellitebased radiance measurements and includes the effect of the terrain topography on the incident solar radiation. The terrain slope and its aspect are generated from the digital elevation model of the region. The analysis carried out over the Beas Basin, Himachal Pradesh, India, indicated a variation of 22% in the amount of incident solar radiation for an increase of 10‡ in terrain slope. Further, the terrain with south-east aspect received maximum amount of solar radiation. The large differences observed between the uncorrected and terrain corrected reflectances in the shortwave infrared band (B5), which is not saturated over the snow covered region, suggest that the terrain slope and its aspect cannot be neglected in estimating the accurate spectral reflectances over the Himalayan terrain.

Nonlinear estimation of spectral reflectance based on Gaussian mixture distribution for color image reproduction.

Nonlinear estimation method of spectral reflectance from camera responses is proposed. The proposed method minimizes the mean square error of spectral reflectance when the reflectance can be regarded as a random sequence of Gaussian mixture distribution. In computer simulations, 168 samples of spectral reflectance from a color chart are estimated from their image signals obtained by three- and six-band cameras. It is confirmed that the proposed method improves the accuracy in comparison with the conventional Wiener estimation method.

Testing LiSK BRDF models over a semi-arid grassland region with visible and near-infrared ATSR-2 and AVHRR data

This paper assesses the capability of the Roujean and LiSparse-MODIS-RossThin linear semi-empirical kernel-driven (LiSK) bidirectional reflectance distribution function (BRDF) models to predict bidirectional reflectance at geometries other than those of the observations used to invert the model, when the models are inverted against a sparse set of angular samples from 21 orbits (3-19 August 1996) of the operational Advanced Very High Resolution Radiometers (AVHRRs) on NOAA TIROS series AM (morning) and PM (evening) satellites. Red ('visible') and near-infrared (NIR) spectral reflectance estimates acquired at 4:40 GMT on 14 August 1996 by the Along-Track Scanning Radiometer-2 (ATSR-2) sensor flown on the European Space Agency's ERS-2 satellite are used as reference data. The test area is a semi-arid grassland region in Inner Mongolia, P.R. China, bounded by 42.84°-44.71° N and 112.40°-116.05° E. The results show that in spite of the difficulties posed by such a task, LiSK models can be inverted against multiangular AVHRR observations to predict bidirectional reflectance at the acquisition geometry of the ATSR-2 with reasonable accuracy: the rms. error of the reflectance predictions made by both models is less than 4% for the nadir views and less than or equal to 6% in the forward views. These error values are less than one-half those provided by a 13 August 1996 AM AVHRR scene in the 0.65 w m channel and about one-seventh of those for the AVHRR scene in the 0.87 w m (NIR) channel, in both nadir and forward views.

An analytical model for the cloud-free atmosphere/ocean system reflectance

An analytical model for the estimation of spectral and directional reflectance over the ocean is developed. The model accounts for the first and second orders of reflection at the ocean surface, and/or scattering by the atmosphere and the water column. The model parameters are the optical thickness and Ångström exponent of aerosols, the spectral albedo of the ocean, and the directional specular reflectance of the ocean surface (glitter), the latter being a function of the wind speed and direction. We show that our analytical model is sufficient to predict all reflectance angular and spectral variations by comparing its results with a Monte Carlo simulation that accounts for the higher orders of scattering. Using this analytical model, actual bidirectional and spectral measurements are easily inverted in order to retrieve geophysical and optical parameters of the ocean-atmosphere system (the application to the data acquired by the Airborne POLDER is developed in a companion article).

Analysis of condensates formed at the viking 2 lander site: The first winter

A thin light-colored ground covering appeared on the surface of Mars near the Viking 2 lander from L s = 230° to L s = 16°, a total of 249 Mars days, during the lander's first winter on the surface. This paper presents a reduction of applicable lander imagery during the period. Imaging sequences, relative surface albedo, spectral reflectance estimates, and limited photometric data are presented and compared with previous laboratory measurements. Photometric data are best fit by an average Minnaert k = 1.1 (blue), k = 1.0 (green), and k = 0.95 (red). Appearance and disappearance rates, spectral reflectance, and photometric data all tend to confirm an earlier proposal that the covering was a combination of H 2O and CO 2, which fell already condensed onto dust particles brought northward by the season's first major dust storm. Under this assumption, the covering thickness is estimated to be between 0.5 and a few millimeters.

1009 ESTIMATION OF SERUM BILIRUBIN BY SPECTRAL REFLECTANCE OF THE SKIN

A study was undertaken to determine whether the serum bilirubin level could be estimated by the light reflectance of skin. The reflectance of skin over the abdomen was measured at five wavelengths (λ = 629, 556, 511, 465, 424), and compared with reflectance of a white standard. A formula was derived which related measurements of skin reflectance with serum bilirubin level. Thirty Black and 14 Caucasian term infants were then prospectively studied with simultaneous spectral reflectance and serum bilirubin levels. All infants were less than one week of age, and none had received phototherapy. Serum bilirubin levels ranged from 1 mg/dl to 12 mg/dl and correlated with spectral reflectance estimates with a coefficient of 0.94. Agreement was within 2 mg/dl in all cases. Measurement of spectral reflectance of skin is an accurate and non-invasive screening method for estimation of serum bilirubin.

Estimation of spectral reflectance curves from multispectral image data

A technique is presented for estimating spectral reflectance curves from multispectral image data even if the spectral samples are obtained from channels whose spectral responsivity is not narrowband. It is demonstrated that these reflectance estimates can be written as a linear combination of the spectral samples and that, analogous to Shannon's sampling theorem, if the spectral reflectance is a natural cubic spline, it can be estimated exactly provided the number of spectral channels is sufficiently large. Simulation results suggest that the accuracy of the spectral reflectance estimates is quite good and very insensitive to the spectral responsivity shapes.

Spectrophotometric and color estimates of the Viking Lander sites

The spectral radiance and color of the Martian sky and soil and the spectral reflectance of soil features are estimated from six-channel (0.4-1.0 micron) spectral data obtained with the Viking lander cameras. Images taken near local noon from the two landers reveal a sky that is brighter near the horizon than the soil but with a similar spectral radiance shape and color. The scenes are predominantly moderate yellowish brown in color with only subtle variations except for some dark grey rocks. Most spectral reflectance estimates are similar: they rise rapidly with increasing wavelength between 0.4 and 0.8 micron and with only a few exceptions exhibit a pronounced minimum centered about 0.93 micron. These characteristics are consistent with an abundance of Fe(3+)-rich weathering products, notably nontronite. However, the delineation of the number and abundances of total mineral phases requires further analyses and laboratory comparisons. Reflectance estimates for rocks have not been repeatable, probably because most rocks have irregular pitted surfaces that introduce significant shadowing components.