Walter Debruyn

A band selection technique for spectral classification

In hyperspectral remote sensing, sensors acquire reflectance values at many different wavelength bands, to cover a complete spectral interval. These measurements are strongly correlated, and no new information might be added when increasing the spectral resolution. Moreover, the higher number of spectral bands increases the complexity of a classification task. Therefore, feature reduction is a crucial step. An alternative would be to choose the required sensor bands settings a priori. In this letter, we introduce a statistical procedure to provide band settings for a specific classification task. The proposed procedure selects wavelength band settings which optimize the separation between the different spectral classes. The method is applicable as a band reduction technique, but it can as well serve the purpose of data interpretation or be an aid in sensor design. Results on a vegetation classification task show an improvement in classification performance over feature selection and other band selection techniques.

Generic wavelet-based hyperspectral classification applied to vegetation stress detection

This communication studies the detection of vegetation stress in hyperspectral data. Compared to traditional vegetation stress indices, the proposed approach uses the complete reflectance spectrum and its wavelet representation. The detection strategy is formulated as a classification problem. Experiments are conducted on fruit tree stress detection. The experiments show the superior performance of the proposed strategy and demonstrate its generic nature.

APEX: current status of the airborne dispersive pushbroom imaging spectrometer

Over the past few years, a joint Swiss/Belgium ESA initiative resulted in a project to build a precursor mission of future spaceborne imaging spectrometers, namely APEX (Airborne Prism Experiment). APEX is designed to be an airborne dispersive pushbroom imaging spectrometer operating in the solar reflected wavelength range between 4000 and 2500 nm. The system is optimized for land applications including limnology, snow, and soil, amongst others. The instrument is optimized with various steps taken to allow for absolute calibrated radiance measurements. This includes the use of a pre- and post-data acquisition internal calibration facility as well as a laboratory calibration and a performance model serving as a stable reference. The instrument is currently in its breadboarding phase, including some new results with respect to detector development and design optimization for imaging spectrometers. In the same APEX framework, a complete processing and archiving facility (PAF) is developed. The PAF not only includes imaging spectrometer data processing up to physical units, but also geometric and atmospheric correction for each scene, as well as calibration data input. The PAF software includes an Internet based web-server and provides interfaces to data users as well as instrument operators and programmers. The software design, the tools and its life cycle are discussed as well.

Wavelet-based feature extraction for hyperspectral vegetation monitoring

The high spectral and high spatial resolution, intrinsic to hyperspectral remote sensing, result in huge quantities of data, which slows down the data processing and can result in a poor performance of classifiers. To improve the classification performance, efficient feature extraction methods are needed. This paper introduces a set of features based on the discrete wavelet transform (DWT). Wavelet coefficients, wavelet energies and wavelet detail histogram features are employed as new features for classification. As a feature reduction procedure, we propose a sequential floating search method. Selection is performed using a cost function based on the estimated probability of error, using the Fisher criterion. This procedure selects the best combination of features. To demonstrate the proposed wavelet features and selection procedure, we apply it to vegetation stress detection. For this application, it is shown that wavelet coefficients outperform spectral reflectance and that the proposed selection procedure outperforms combining the best single features.

Nitrous oxide emissions from waste water

The estimation of nitrous oxide emissions is complicated by the high degree of uncertainty on the emission factors involved and by the limited acquaintance with all significant nitrous oxide sources. A potentially important source for which emission data are lacking is the sewage system transporting waste water from human activities. For this study an experimental measurement campaign has been carried out on waste water sampled at different sewage treatment plants. The nitrous oxide developing from the water samples was monitored by means of gas chromatography. The methodological analysis was based on the concentration/time curves obtained. Our results indicate that the formation of nitrous oxide from the waste water matrices results from microbiological denitrification. We deduced tentative emission factors for the waste water types studied.

Retrieval of oceanic constituents from ocean color using simulated annealing

Abstract : The color of the sea is determined by the contents of the water, especially the concentrations of suspended particulate matter (SPM), phytoplankton pigments such as chlorophyll (CHL) and colored dissolved organic matter (CDOM). Reversely, optical sensors that measure the water-leaving reflectance spectra allow us to calculate the desired concentration products. In this paper, a method is introduced that is valid for both case 1 and 2 waters. To this end, model is fitted to reflectance spectra, using simulated annealing for optimizing the mean square of the reflectance over all spectra.

The measurement of nitrous oxide emissions from sewage systems in Belgium

To quantify the nitrous oxide emissions from waste water, an experimental measurement campaign has been set up; waste water was sampled at the collector tubes entering sewage treatment plants and at the settling tanks in these plants. The gas phase developing in the static head space of the water samples was analysed; gas chromatography by means of electron capture detection was the analytical tool by which the nitrous oxide concentration in batch samples of gas was determined. The methodological analysis was based on the concentration/time curves obtained.The formation of nitrous oxide from the waste water matrices is the result of the microbiological denitrification of the organic substrate present; this could be deduced from the response of the nitrous oxide signal to the addition of NaNO3, NH4NO3 and (NH4)2SO4 to the samples. Application of the Lineweaver-Burk kinetic equation for enzyme-catalysed reactions on our results, combined with the yearly mean nitrate concentration and the seasonal mean waste water temperature, enabled us to deduce emission coefficients for the two types of waste water sampled: raw waste water: (4.3 ± 1.0)µg N2O/gss, settled waste water: (800 ± 180)µg N2O/gss, where “gss” stands for “gram suspended solids”, a water quality parameter continuously monitored in Belgium.

Classifying hyperspectral airborne imagery for vegetation survey along coastlines

This paper studies the potential of airborne hyperspectral imagery for classifying vegetation along the Belgian coastlines. Here, the aim is to build vegetation maps using automatic classification. Besides a general linear multiclass classifier (Linear Discriminant Analysis), several strategies for combining binary classifiers are proposed: one based on a hierarchical decision tree, one based on the Hamming distance between the codewords obtained by binary classifiers and one based on the coupling of posterior probabilities. In addition, a new procedure is proposed for spatial classification smoothing. This procedure takes into account spatial information by letting the decision for classification of a pixel depend on the classification probabilities of neighboring pixels. This is shown to render smoother classification images.

Status of the airborne dispersive pushbroom imaging spectrometer APEX (Airborne Prism Experiment)

Over the past few years, a joint Swiss/Belgian initiative resulted in a project to build a new generation airborne imaging spectrometer, namely APEX (Airborne Prism Experiment) under the ESA funding scheme named PRODEX. APEX is designed to be a dispersive pushbroom imaging spectrometer operating in the solar reflected wavelength range between 400 and 2500 nm. The spectral resolution is designed to be better than 10 nm in the SWIR and 5 nm in VIS/NIR range of the spectrum. The total FOV is on the order of /spl plusmn/14/spl deg/, recording 1000 pixels across track, and max. 300 spectral bands simultaneously. The final radiance data products are well characterized and calibrated to be traceable to absolute standards. APEX is subdivided into an industrial team responsible for the optical instrument, the calibration home base, and the detectors, and a science and operational team, responsible for the processing and archiving of the imaging spectrometer data, as well as its operation. APEX is in its design phase with partial breadboarding activities and will be operationally available to the user community in the year 2005.

Hyperspectral classification applied to the Belgian coastline

Hyperspectral image classification impose challenging requirements to a classifier. It is well known that more spectral bands can be difficult to process and introduce problems such as the Hughes phenomenon. Nevertheless, user requirements are very demanding, as expectations grow with the available number of spectral bands: subtle differences in a large number of classes must be distinguished. As multiclass classifiers become rather complex for a large number of classes, a combination of binary classification results are often used to come to a class decision. In this approach, the posterior probability is retained for each of the binary classifiers. From these, a combined posterior probability for the multiclass case is obtained. The proposed technique is applied to map the highly diverse Belgian coastline. In total, 17 vegetation types are defined. Additionally, bare soil, shadow, water and urban area are also classified. The posterior probabilities are used for unmixing. This is demonstrated for 4 classes: bare soil and 3 vegetation classes. Results are very promosing, outperforming other approaches such as linear unmixing.