Let {X1(t1),t1≥0}, …,{Xp(tp),tp≥0} be p independent stable component processes in Rd. We obtain the law of the iterated logarithm for the measure on the product set ∏ i=1pXi[0,1]. By applying the density theorem, we derive the exact packing measure of the product set. As a special case, we deduce the exact packing measure of the product set of ranges and graphs for p independent stable processes in Rd.
The complexity of the dual basis of a type-Ⅰ optimal normal basis of Fqn over Fq was determined to be 3n-3 or 3n-2 according as q is even or odd, respectively, by Wan and Zhou in 2007. We give a new proof to this result by clearly deriving the dual of a type-Ⅰ optimal normal basis with the aid of a lemma on the dual of a polynomial basis.
Let B(X) and B(Y) be the Banach algebras of all bounded linear operators on infinite dimensional complex Banach spaces X and Y, respectively. It is proved that an additive surjective map φ from B(X) to B(Y) preserves the nullity in both directions if and only if there exists invertible bounded linear or conjugate linear operators U:X→Y and V:Y→X, such that φ(T)=UTV for all T∈B(X). A form for additive surjective maps preserving the deficiency in both directions is also obtained.
In this paper, multiwavelength data of HPLC-DAD were used to discriminate different brands of blue ballpoint pen ink. Dyes from inks of sixty-eight blue ballpoint pens were analyzed by using HPLC-DAD. Dye components detected in the inks were further identified by using HPLC-mass spectrometry (HPLC-MS). All the inks were firstly classified into nine groups according to the presence or absence of the significant chromatographic peaks. Then the inks in the same group were further discriminated by using PCA with multiwavelength chromatographic data. The results indicated that chromatographic data extracted from three wavelengths were more potent in the discrimination as compared with a single wavelength chromatographic data. By combining the two discrimination steps, we can discriminate about 94% of blue ballpoint pens according to the brands.
Polder is an important geographic unit of the water net plain areas in southern China, and study of the polder water ecosystem service functions has important significance. We selected polders in western part of Taihu basin. On the basis of the available polder data, supplying, regulating, and maintaining functions of ecosystem were assessed quantitatively. The results show the following conclusions. 1)The total quantities of water ecosystem service functions, as well as the amount of per unit area, are different in different spaces. 2) The management and recovery of ecosystem service functions in polders should refer to the results of assessment and be suited to the local conditions. 3)The management of ecosystem service function in polders should have a unified plann and harmonie management for determination of more rational land use and mode of production in polders.
Neural network is a good modeling method in water color remote sensing field, and it needs a large amounts of training data sets. However, in situ data sets are far from enough for meeting the requirement of large amounts of training data sets in neural network. So Hydrolight is used to cover the shortage, and it can simulate large-scale data sets. Based on in situ data of 7 months from 2008 to 2011 in Taihu Lake and by using related numerical statistical methods, the primary required simulated parameters such as the mass-specific absorption coefficients of chlorophyll particles, mineral particles, and colored dissolved organic matter (CDOM) and the mass-specific scattering coefficients of chlorophyll particles and mineral particles are determined. Results show that, compared with the measured remote sensing reflectance spectrum, the simulated remote sensing reflectance spectrum matches well with the former one and the correlation coefficient reaches 0.95 generally. This study provides a good foundation for building large scale training data sets of Taihu Lake in neural network simulation.
Landsat 5 TM and HJ-1 CCD images in 2009 were selected for land-cover classification in mountainous region of southern Anhui. Combining with field survey data, we analyzed land-cover characteristics to dig out spectrum, texture, spatial and temporal information, and obtained the feature bands of the corresponding land-cover types. Using See 5.0 software, we oltained the optimal feature bands and thresholds of land-cover types based on training samples. The rule sets of land-cover classification were established using eCognition Developer 8.64 software. The results showed that, by using the object-oriented classification method based on multi-temporal remote sensing images, the land-cover classification with accuracies higher than 92% was acquired. This method is suitable for land-cover classification in the mountainous region of southern Anhui.
For truly describing the process and trend of land-use change, HJ-CCD and ETM imagery was used to obtain land-use change information in Changsha in the recent ten years(2000-2010). Then we used models including land-use dynamic model, degree change model, and transition matrix to quantitatively analyze the land-use change process information in Changsha in the ten years. Furthermore, Ps trend index was used to forecast the trend for each type of land-use changes in future. The results are the following. The spatial transfer of the artificial surface is not frequent, and the area will be rapidly expanded due to transfer from farmland. The spatial transfers of farmland and woodland are frequent. Most of farmlands will become artificial surfaces in future. Forestland will transfer most likely to artificial surfaces and also to grass. Grass will expand due to transfer from badland. Badland will continuously become artificial surfaces.
Selection of algorithms for forest fire spreading simulation is important. In this paper, the traditional boundary interpolation algorithm is optimized and then compared with the maze algorithm in forest fire spreading simulation. The results are the following. The maze algorithm consumes more time (204、458、1325 s) than the boundary interpolation algorithm (5、7、16 s). The time consumed by the maze algorithm will doubly increase;As far as accuracy of simulation results concerned, the boundary interpolation algorithm has a lower accuray(59.75%) than the maze algorithm, but it shows simple changes of the forest fire spreading generally. Results obtained from maze algorithm provide more detailed information for the forest fire spreading and it resembles closely to actual spreading area.
Ecological footprints (EF) and ecological capacity (EC) in Liangshan prefecture from 2004 to 2011 were studied by using the ecological footprints model. Ecological footprints and ecological capacity in Liangshan prefecture in future were predicted by using the growth prediction model. Results showed that in Liangshan prefecture the per capital EF increased from 0.506 66 hm2 of 2004 to 0.593 46 hm2 of 2011; the per capital EC increased from 0.873 43 hm2 of 2004 to 0.717 70 hm2 of 2011; and the ecological surplus constantly decreased. The ecological surplus will turn into ecological deficit in 2015, and the ecological deficit will constantly grow. The efforts for realization of sustainable development in Liangshan prefecture are discussed.
Based on economic environmental factor, social environmental factor, and ecological environmental factor, we constructed a comprehensive evaluation index system on support of industrial developing environments. We find that the correlation between industrial developing environments and industrialization shows a similar periodicity to industrialization evolution. The correlation between economic environment and industrialization level is high, which shows that the economy is still the strong driving force for the industrial development. Although the correlation with ecological environmental factor is small, the rising trend reflects that the degree of attention to the ecological environment is increasing in the whole process.
Our study aims at assessment of sustainable development of 30 municipalities and provincial capitals in China. First, we designed an assessment indicator system, which is composed of economic, social, and environmental subsystems. Secondly, we used the entropy method to calculate each indicator's weight and then applied grey relational analysis to evaluate sustainable development capacities of these cities. The results show that, although cities in the east generally have greater sustainable development capacities, the developments among the three subsystems are uneven. Some cities in central and western China are well-developed in terms of their economic and social systems. However, the pollutant emission intensities of most cities in the west are relatively high, which may lead to poor performance in environmental sustainability.
In this study,given shape templates of some object,we model them by using kernel principal component analysis and then propose a new object extraction method for high resolution remote sensing images, which integrates shape prior and several image appearance information,including edge,color, and shadow. Based on active contour model,a new energy function is constructed and minimized through an iterative global optimization method to get the accurate segmentation results. Experimental results show that our method has high efficiency, high accuracy, and the robustness with respect to various background disturbances.
In order to meet the requirement of terrain mapping in real time, we solve the problem of large computation in algorithm and provide real-time DEM generating algorithm based on airborne dual-antenna InSAR system. The algorithm utilizes non-linear ECS auto-registration imaging algorithm to generate high-accuracy registrated dual-channel SAR images, and omits the sabsequent time-consuming registration course. Fast filter is performed on the two SAR images to generate high-quality interferometric fringe, and then fast phase unwrapping algorithm based on partitioning the dense residues area is done to get the unwrapped phase. For the traditional time-consuming calibration problem, estimation of initial phase deviation in real time based on external coarse precision DEM data is performed. By combining the parameters such as baseline length, baseline obliquity, and absolute time delay, DEM is generated in real time. Through analyzing the real-time performance of the algorithm, the hardware structure to realize the algorithm is brought forward. The processed results of the real InSAR data show the validation of the algorithm.
We propose a QoS-driven resource allocation algorithm in the downlink orthogonal frequency division multiple access (OFDMA) network by integrating information theory with the concept of effective capacity (EC). By exploring the decoupling property of real-time session,we maximizes the network EC by increasing the EC of file transfer users gradually until the system power reaches the threshold while keeping the EC of real-time session users unchanged. Simulation shows that the proposed algorithm significantly improves the overall EC under the delay constraint.
An adversary can infiltrate online social networks (OSNs) on a large scale by deploying socialbot network, which is an army of socialbot accounts. This will endanger the information security of online social network and users. To solve the problem, we propose a detection method based on the population characteristics. We extract the following population characteristics: centralized created time, similar screen names, and coincident active time. On the basis of the extracted charateristics and by using date mining method, the method is proposed to detect socialbots networks. The method is used in a data set of 480 000 users of sina microblog and detects many socialbots networks which include 6 899 socialbots accounts. The low false negative rate and false positive rate indicate that the method is feasible and effective.
To clearly describe the change of decision indicator weights,we propose a new weight-assessing model based on the behavior mechanism and the habitual domain theory and we prove that the index weight relates to the control coefficients and the degree of association.The process of weighting is given. The model has been applied in the safety management decision of a coal mining enterprise. The results show the applicability of this new model.
HJ-1C is the first civil S-band synthetic aperture radar(SAR) satellites launched by China, and the equipped SAR has the traits of all-weather, all-time, and penetrability. We focus on HJ-1C satellite. As SAR satellites have complex imaging algorithm and need large amount of calculation, we propose GPU-based SAR real-time imaging system based on GPU-based parallel processing technology. The results show that the processing speed of the SAR real-time imaging system is higher than 600 Mbps, meeting the real-time imaging processing requirements for practical SAR satellite ground receiving system.
In speaker recognition, the selection of useful information is an important pre-processing step. Usual ways for selection of the useful information are based on energy. However, between useful information and energy there are no necessary connections. After analying the traditional selection ways, we propose a phoneme decoder based data selection algorithm. Through analysis of the phoneme recognition results, all vowels are kept and some useless consonants are filtered. The speaker recognition experiment results show that the proposed method is superior to the traditional energy-based data selection algorithms such as G.723.1 algorithm and the recently proposed cross entropy based order statistics filtering algorithm.