This paper proposes an automatic framework for land cover classification. In majority of published work by various researchers so far, most of the methods need manually mark the label of land cover types. In the proposed framework, all the information, like land cover types and their features, is defined as prior knowledge achieved from land use maps, topographic data, texture data, vegetation’s growth cycle and field data. The land cover classification is treated as an automatically supervised learning procedure, which can be divided into automatic sample selection and fuzzy supervised classification. Once a series of features were extracted from multi-source datasets, spectral matching method is used to determine the degrees of membership of auto-selected pixels, which indicates the probability of the pixel to be distinguished as a specific land cover type. In order to make full use of this probability, a fuzzy support vector machine (SVM) classification method is used to handle samples with membership degrees. This method is applied to Landsat Thematic Mapper (TM) data of two areas located in Northern China. The automatic classification results are compared with visual interpretation. Experimental results show that the proposed method classifies the remote sensing data with a competitive and stable accuracy, and demonstrate that an objective land cover classification result is achievable by combining several advanced machine learning methods. 相似文献
Ray representation of electromagnetic resonance (reverberation) mode in subsurface circular cylindrical dielectric resonator (including void) is proposed. The modal ray path must be regular polygon or polystar. Travel time formulas for fundamental resonance multiples and the detecting conditions for non-exact-backscattering rays are derived. Simulation of time–distance curves of a modal hollow concrete block is generally concurred to the measured ground penetrating radar signal pattern. The proposed modeling method can be generalized to other resonant cavities with different profiles and provide a sound base for further applications of other more complicated geophysics science and engineering fields, particularly in ray-based tomography. 相似文献
The Tarim Desert Highway in Xinjiang, China, the longest one in the world, has a length of 562 km, about 80% of which runs across, from north to south, the Taklimakan Desert. Obviously, the main problem of the road maintenance is the blown sand disaster. The research results showed: (1) the physical environment along the desert highway is characterized by strong winds, fine and loose ground materials, different dunes and so on, which provides the dynamical condition and material source for the formation of blown sand disaster to the road and its shelter system. Meanwhile, the trend and cross-section of the road and the structure of the shelter system, as damage objects, play important roles in the formation process of blown sand disaster; (2) the blown sand disaster to the shelter system is original from the intrusion of the drift sands and mobile dunes outside the shelter system, and the wind erosion and sand deposit caused by the air stream changes on the ground in the shelter system. The main damage object in the Tarim Desert Highway is the shelter system presently. The damage forms include wind erosion, sand burying and dune covering; and (3) the damaged length of the blocking sand fences is 83.7%, 88.4%, 72.4%, 72.8% and 40.3% and the damaged area of the straw checkerboard belts is 73.1%, 58.2%, 44.5%, 35.4% and 36.6%, in turn, in 5 different landform units from north to south, and, the disasters to fences and the straw checkerboard belts are 79.5% and 57.6% in the compound dunes while they are 64.6% and 37.7% in the interdunes respectively.
Spectrum analyses of water quality time series have been carried out for five hydrometric stations including Wuhan hydrometric
station of the Changjiang(Yangtze) River, etc. The fluctuations of Ca2+, Mg2+ and HCO
3−
concentrations in river water under different physical geography conditions have about two-year cycle which is corresponding
to hydrometeorological quasi-biannual-oscillation (QBO). Na+, Cl− SO
42−
have about two-year cycle in the area lightly affected by human activities while two-year cycle doesn’t exist in the area
heavily affected by human activities. All the fluctuations of major ions have about three-month cycle. The river diseharge
fluctuation accounts for 43.9%, 45.1%, 54.3%, 33.9%, 30.3% and 42.7% of the variance of Ca2+, Mg2+, HCO
3−
, Na+, Cl−, SO
42−
, respectively, at Wuhan from 1962 to 1985. According to the spectrum characteristic of major ions, the duration of the time
series has to be at least 13 years for trend analysis of monthly water quality data.
Foundation item: Under the auspices of the National Natural Science Foundation of China (No. 49671017).
Biography: XIA Xing-hui(1971-), female, a native of Hunan Province, Ph. D. Her research interest includes environmental chemistry. 相似文献