A new method of extracting broad-leaf weeds from grassland, Variance Extraction Method (VEM), has been proposed. The method applies texture analysis to weed extraction. Statistical variance was found suitable as a feature parameter of the texture analysis. Only weeds can be extracted because variance calculated on the broad leaves of the weeds is small and it is large in the grassland region. In this paper, characteristics of the proposed method are closely investigated and a conditional equation of extraction is presented for the design of weed system.
A supervised classification method using a self-organizing map (SOM) is proposed to classify remote sensing data. SOM has a characteristic that a probability density function of input data is represented as a feature map. The proposed method is realized by creating a category map from the feature map of SOM. The category map can visualize characteristics of SPOT HRV data and it is also employed as a supervised classification method. The proposed method extracts liquefied area in Kobe (Japan) damaged by the 1995 Hyogoken Nanbu earthquake using the SPOT HRV data and the category map. As an experimental result, it is shown that classification accuracies of the proposed method are higher than those of the maximum likelihood and the back-propagation methods.
The Yao River basin, located in the Nan Province, north Thailand, is one of the most critical watersheds in terms of soil erosion. USLE method was combined with GIS technology so as to estimate the soil erosion volume in this basin. As a result, the total amount of 131, 500 ton/year for the whole basin was calculated. Based on the spatial distribution of USLE value, the most hazardous sub basin with 20, 075 ton/year was detected on the left side of the lower Yao River. Then, USLE analysis at larger scale was conducted for this sub-basin. By applying the higher resolution DEM and aerial photo interpretation, it became clear that the actual value was 8, 940 ton/year because the slopes were less steep and more vegetated. Although there is no land conservation practice, the supposed three conditions (contouring, stripping and terracing) were simulated in order to evaluate the effectiveness of each type of practice. The result shows that the erosion volume was reduced to less than half by constructing the stripping or terracing practice.
A field campaign for various calibration of ADEOS/AVNIR and OCTS was conducted at Ivanpah Playa, California, USA on March 4 1997. Through a careful analysis of the observed optical depth of the atmosphere, the measured surface Bidirectional Reflectance Distribution Function (BRDF) and the other meteorological data, Top of Atmosphere (TOA) radiance was estimated based on the Radiative Transfer Equation (RTE) and was compared to the real AVNIR and OCTS derived radiance. It is found that the difference between both ranges from around -3.5 to -20% for AVNIR while those for OCTS ranges from approximately -9.5 to -15%.