Integrated Data Fusion Research Articles

A Novel Remote Sensing Image Classification Scheme Based on Data Fusion, Multiple Features and Ensemble Learning

In this paper, we propose a novel scheme to improve the accuracy of remote sensing image classification by integrating data fusion, multiple feature combination and ensemble learning. Intensity-Hue-Saturation (IHS), Gram-Schmidt (GS), Brovey and wavelet fusion methods are first performed to obtain the optimal fusion images of high resolution and multispectral images. Support Vector Machine (SVM) classifier is then adopted to classify the fused image with different feature sets, and ensemble learning algorithm based on dynamic classifier selection (DCS) is finally used to integrate multiple classification maps. The proposed classification scheme is implemented with three remote sensing data sets, obtaining the highest overall accuracy and kappa coefficient in all cases (92.63% and 0.8917 for BJ-1 data set, 81.89% and 0.7513 for Landsat TM and SPOT4 data set, 92.21% and 0.8838 for ALOS data set respectively). The experimental results show that the integration of data fusion, feature combination and ensemble learning improves the classification performance obviously and has great potential in practical uses.

WITHDRAWN: Advanced condition monitoring of tapered roller bearings, Part II: With integrated data fusion techniques

WITHDRAWN: Advanced condition monitoring of tapered roller bearings, Part II: With integrated data fusion techniques

Identifying Unsaturated Hydraulic Parameters Using an Integrated Data Fusion Approach on Cross‐Borehole Geophysical Data

Cross‐borehole geophysical data can provide valuable information concerning hydrologic properties of the unsaturated zone. Such data are most often used sequentially, where images of soil physical properties are obtained through numerical inversion and then converted to hydrologic state properties using petrophysical relationships. If not accounted for, inversion artifacts are transferred to the resulting hydrologic images. We propose a framework in which multiple geophysical data sets can be incorporated using an integrated data fusion approach. The geophysical data collected are integrated in a forward modeling approach to evaluate a series of plausible hydrologic models. The approach permits an evaluation of the sensitivity of geophysical data for constraining hydrologic model parameters. We illustrate the approach using geophysical data collected during a dual water and solute tracer experiment in the unsaturated zone. Cross‐borehole ground penetrating radar and electrical resistivity tomography, measuring electromagnetic travel time and electrical transfer resistances, respectively, were collected during a 20‐d period. As a first approximation, one‐dimensional flow was considered and three models (one, two, and five layers) of the subsurface were evaluated. The five‐layered model was found to be the only model capable of mimicking the infiltration pattern satisfactorily. The results showed that only the hydraulic conductivity and one of the parameters (empirical parameter n) describing the soil moisture release curve for three of the five layers could be constrained by the data, illustrating the nonuniqueness of the problem. The data fusion approach proved, however, that application of multiple geophysical methods may reduce hydraulic parameter uncertainty.

Structural Damage Detection by Integrating Data Fusion and Probabilistic Neural Network

Over the past two decades, multi-sensor data fusion method has attracted increasing attention to structural health monitoring due to its inherent capabilities in extracting information from different sources and integrating them into a consistent, accurate and intelligible data set. Meanwhile, since the probabilistic neural network (PNN) describes measurement data in a Bayesian probabilistic approach, it has been successfully applied to structural damage detection (Jiang et al. 2004; Klenke et al. 1996; Ko et al. 1999; Ni et al. 2001). In order to make full use of multi-sensor data (or information) from multi-resource and to improve the diagnosis accuracy of the health conditions for complex structures, it is advisable to combine these methods and exploit their individual advantages. In this paper, a 5-phase complex structural damage detection method by integrating data fusion and PNN is developed and implemented. The proposed method is then applied to damage detection and identification of two simulation examples. The result shows that the proposed method is feasible and effective for damage identification.

Data Fusion and Decision Support for Command and Control

The problems and benefits of integrated data fusion and decision support are discussed across the spectrum of antisubmarine, tactical air, and land warfare. Then a general architecture of activities is proposed. This architecture includes data association, multitarget tracking, data combination, sensor management, situation assessment, and alternative analysis.