Despite recent rapid advancement in remote sensing technology, accurate mapping of the urban landscape in China still faces a great challenge due to unusually high spectral complexity in many big cities. Much of this complication comes from severe spectral confusion of impervious surfaces with polluted water bodies and bright bare soils. This paper proposes a two-step land cover decomposition method, which combines optical and thermal spectra from different seasons to cope with the issue of urban spectral complexity. First, a linear spectral mixture analysis was employed to generate fraction images for three preliminary endmembers (high albedo, low albedo, and vegetation). Seasonal change analysis on land surface temperature induced from thermal infrared spectra and coarse component fractions obtained from the first step was then used to reduce the confusion between impervious surfaces and nonimpervious materials. This method was tested with two-date Landsat multispectral data in Shanghai, one of China’s megacities. The results showed that the method was capable of consistently estimating impervious surfaces in highly complex urban environments with an accuracy of R2 greater than 0.70 and both root mean square error and mean average error less than 0.20 for all test sites. This strategy seemed very promising for landscape mapping of complex urban areas.
Applications of the intensity-hue-saturation (IHS) based image fusion techniques in resource inventory and environmental monitoring are usually hampered by considerable spectral distortion in the spatially enhanced image. The image transform model needs to be regulated via a proper design of the weight structure and controlling parameters to achieve a better spectral fidelity. Use of localized weight estimation and an output constraint to modify the generalized intensity-hue-saturation transform (GIHS) for the purpose of rectifying digital numbers of the fused image back to their original counterparts are proposed. The weight localization was achieved via land cover classification of the multispectral data, and the spectral constraint was constructed using a ratio between individual spectral bands stratified with each land cover type and the modified image intensity value. This method was compared both spatially and spectrally with the traditional IHS and the GIHS that has a weight structure induced from the sensor's spectral response characteristics. Experiments with WorldView-2 multispectral and panchromatic data indicated that the new image fusion approach achieved the highest level of spectral fidelity with enhancement of spatial details comparable to the other IHS-based methods.
Hyperspectral imagery has been widely used in terrain classification for its high resolution. Urban vegetation, known as an essential part of the urban ecosystem, can be difficult to discern due to high similarity of spectral signatures among some land-cover classes. In this paper, we investigate a hybrid approach of the genetic-algorithm tuned fuzzy support vector machine (GA-FSVM) technique and apply it to urban vegetation classification from aerial hyperspectral urban imagery. The approach adopts the genetic algorithm to optimize parameters of support vector machine, and employs the K-nearest neighbor algorithm to calculate the membership function for each fuzzy parameter, aiming to reduce the effects of the isolated and noisy samples. Test data come from push-broom hyperspectral imager (PHI) hyperspectral remote sensing image which partially covers a corner of the Shanghai World Exposition Park, while PHI is a hyper-spectral sensor developed by Shanghai Institute of Technical Physics. Experimental results show the GA-FSVM model generates overall accuracy of 71.2%, outperforming the maximum likelihood classifier with 49.4% accuracy and the artificial neural network method with 60.8% accuracy. It indicates GA-FSVM is a promising model for vegetation classification from hyperspectral urban data, and has good advantage in the application of classification involving abundant mixed pixels and small samples problem.
The method of "subpixel analysis" is explored to decompose mixed pixels of mangrove for biomass quantification. The
basic idea is to treat the non-mangrove spectra of a mixed pixel as the background noise and iteratively remove them
from further processing, so that the residual radiance can be matched to the characteristics of, and labeled as, the sampled
mangrove spectra. This method requires spectral training only on the targeted cover type (i.e. mangrove in this study),
thus it may drastically reduce the amount of human interference and minimize subjective bias in the analytic process. In
addition, it can deal with complex and diverse spectra of the same target for better results. A DigitalGlobe's Quickbird
multispectral image of Beilun Estuary was used as a test dataset to demonstrate this approach, with mangrove cover of
the region being quantified into eight standardized biomass levels. The verification of the model results was performed
using Quickbird panchromatic data from the same acquisition. An overall accuracy of 86.1% (Kappa=0.844) was reached,
demonstrating the application potential of the subpixel analysis method in the forest ecosystem research and
management.
This article examines several major bathymetry mapping methods and describes an experimental procedure of
determining seabed bathymetry from multi-source passive remote sensing data. Issues to be addressed include how to
deal with less desirable spectral data quality and the absence of in-situ water depth measurements. A case study was
presented using DigitalGlobe QuickBird and Landsat-7 ETM+ multispectral images of different dates and spatial
resolutions to determine water depth for the Beilun Estuary, China. The preliminary results have led to three findings.
First, it was feasible to use the tidal water line derived from the near-infrared bands as a good approximation of water
surface when observed tidal data is absent. Second, the reflectance ratio transform model developed by Stumpf et al.
was proven suitable for spectrally-based water depth estimation when in-situ data is absent. Finally, the data quality
problem caused by thin clouds could be effectively removed by fusing remote sensing images of two different sources.
This paper presents an unsuccessful attempt to identify different mangrove species from the DigitalGlobe's QuickBird high-resolution multispectral image data for a coastal estuary located in the north of South China Sea. A conventional supervised classification was conducted with 102 signatures trained for five cover classes, with 32 of the signatures being used to separate up to five mangrove species. The results indicated that spectral characteristics alone as provided by the QuickBird's four spectral bands were not sufficient for the discrimination among mangrove species, other information such as textual and structural characteristics of mangrove species would be needed to enhance the discrimination power. In addition, the confusion between upland forests and mangroves render a removal of uplands from the classification process. Finally, the shadow effect within the mangrove patches suggested the use of NDVI in the future classification attempts.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.