Naoko Kosaka

Forest type classification using data fusion of multispectral and panchromatic high-resolution satellite imageries

This paper proposes fusion analysis of high-resolution multispectral and panchromatic satellite imageries for forest type classification. We have shown the performance of forest type classification using panchromatic and multispectral high-resolution QuickBird satellite imageries separately. With texture features obtained from a panchromatic imagery, forest was classified into two types, such as coniferous and broad-leaved forests. On the other hand, with spectral features obtained from a multispectral imagery, forest was classified into six types, such as three coniferous, one broad-leaved and two mixed forests. These results showed that both texture and spectral features are effective for classification of forest types. In this paper, we apply the object-based classification using the common segments obtained from a pansharpen imagery to fusion and single imagery analysis in order to compare the difference only between texture and spectral features. The mean value of each texture and spectral feature from a segment is adopted in the supervised classification, the standard nearest neighbor method, using radiometrically corrected satellite imageries. We selected the contrast as texture feature, and normalized band values and differences between normalized band values as spectral features. From the comparison of the result with ones obtained from a single imagery analysis, we demonstrated that data fusion analysis exceeds a single imagery analysis in accuracy.

ICA-aided mixed-pixel analysis of hyperspectral data in agricultural land

This letter proposes an independent component analysis (ICA)-aided mixed-pixel analysis of periodically distributed hyperspectral data in agricultural land. This method simultaneously estimates the pure spectra and coverage of endmembers, such as crop and soil, from mixed-pixel data which is inevitably included in images observed from high-altitude sensors. The method is effective for agricultural management because the change of observed mixed-pixel data is distinguished into a qualitative spectral one, due to chlorophyll quantity or crop variety, and the quantitative coverage due to growth stages. This method introduces a priori knowledge which is independent of the type of crop and effective in deriving a scaling factor for the independent component (IC), estimated from the ICA process. The fundamental investigation, using hyperspectral data obtained from a crane and an aircraft, shows the applicability of the method.

Salt-damaged paddy fields analyses using high-spatial-resolution hyperspectral imaging system

In agricultural fields, the damage caused by salinized winds is crucial for crops. In order to minimize the damage, it is required to detect the damaged areas and enact the proper procedures in order to save the damaged fields immediately after the disaster. In this paper, we propose indices that can indicate the degree of salt-breezed damages in the early withering-up stage. To detect the indices, high-spatial-resolution hyperspectral data taken in actual damaged paddy fields are analyzed. In addition, the sequential change of hyperspectral data in rice within artificial withering-up experiments is recorded to interpret the fundamental mechanism of the indices. The applicability of the indices for satellite data is also shown by applying them to simulated satellite data. Keywords-component; spectroscopy; rice paddy; hyperspectral; salt-damaged; SPOT5; NDVI; NDGI; withering up;

Analysis of salt-damaged paddy field using SPOT5 satellite images in Yamagata Prefecture

Abstract —In this paper we propose a new method for estimating the degree of salt-damage over a widespread area using SPOT5 satellite images. In the method, the paddy field pixel data are normalized and plotted in the spectral feature space of NDVI (Normalized Differential Vegetation Index) and NDGI (Normalized Differential Green Index). In the feature space, normal paddies clustered along a standard line, whereas the damaged paddies, at an early stage, shifts at the under side of the standard line. Thus the degree of early-stage damage can be detected from the difference from the standard line. The resultant degree of damages are shown in color and plotted on the map. In order to comparatively evaluate our method, we also examined other methods using reflectance of NIR or NDVI. In the comparison, we confirmed that the newly proposed method showed the best match to the ground truth observation. Subsequently, we attempted to apply the method to the images taken after “Tsunami” in Sumatera for testing the availability.

Neural networks for cerebral diagnosis using PET and SPECT

In the interpretation of cerebral functional images such as PET (positron emission tomography) and SPECT (single photon emission CT), well-trained neural networks provide effectively diagnostic information. In this paper we discuss how to make use of a priori knowledge in the use of neural networks for the diagnostic analysis of cerebral functional images.

Leaf area index estimation from hyperspectral data using group division method

In this study, optical remote sensing data were used for leaf area index (LAI) estimation. The LAI is an important measure to increase the yield and adjust the quantity of manure. LAI extracted from remotely sensed data may contribute to grasp the yield of rice at an early stage. Therefore, the purpose of this study is to estimate the LAI through remote sensing. For the purpose of our work, we proposed a Group Division Method. The method can decide a set of optical bands for estimating the value of LAI. The index is extracted by comparing the order of ground truth data with that of spectral data. As a result, the effective index to estimate LAI is made by reflectance relations in 545nm, 1170nm and 1290nm from the hyper-spectral data of rice field in Sakata City, Yamagata prefecture. Furthermore, we applied the index to the data set obtained in Furukawa, Miyagi prefecture to verify the effectiveness of the method. Finally the “LAI estimate map” was made and examined whether this study helped an automatic LAI estimate in wide area.

Monitoring of Bacterial Pustule on Soybean by Neural Network Using Hyperspectral Data

This paper proposes a technique for estimating the soybean bacterial pustule infection using a neural network, provided with hyperspectral data obtained from the cranemounted hyperspectral sensor system. This technique uses the concept of a quotient set in manifold to reduce the feature space of hyperspectral data. This preliminary study shows that the manifold-embedded neural network is effective to classify bacterial pustule degrees in the nonlinearly reduced hyperspectral feature space. Keywords-bacterial pastule; feature map; hyperspectral data; manifold; neural network; quotient set; soybean

Synthesis and analysis of periodically distributed satellite image components by independent component analysis

In analyzing multispectral satellite images, the existence of mixels caused by high altitude observation makes it difficult to distinguish pure spectra from the coverage. In this paper, we firstly propose an analytic method which enable us to separate the change of vegetation into qualitative one due to biological characteristics such as the chlorophyll quantity, and the coverage quantity by utilizing the characteristic feature of the spatial distribution of crops, i.e. periodicity, for independent component analysis (ICA). As the second proposal, we introduce a synthetic method which aims at the estimation of the spatial frequency and direction of periodically distributed pattern from noisy and diffused images by synthesizing the data based on the combination of multi-directional scanning, multiple bandpass filtering and ICA.