Time filter

Source Type

Yang W.,Beijing Forestry University | Qi W.,Beijing Twenty first Century Science and Technology Development Co. | Wang M.,Beijing Normal University | Zhang J.,Beijing Forestry University | Zhang Y.,Beijing Forestry University
Geomorphology | Year: 2017

Major earthquakes in mountain regions have persistent and significant influences on post-seismic landslides but their details have not been well understood. This work uses multi-year high-resolution satellite images and terrain parameters, such as elevation, slope, and aspect, to examine the topographic changes of post-seismic landslides. Despite disturbances during rainy months, landslide areas decreased significantly from 2008 to 2013 in all terrain parameters, indicating that landslide activity near the epicentre has been recovering to the pre-seismic level. The emergence of an increasingly active landslide type shows that landslide debris has been moving from hillslopes to valleys, which could impact post-seismic debris flows. The findings of this work provide important information for post-seismic infrastructure re-construction and disaster prevention in future mountain earthquake events. © 2016 Elsevier B.V.


Wang X.,Beijing Twenty First Century Science and Technology Development Co. | Wang X.,Heilongjiang University | Wu S.,Beijing Twenty First Century Science and Technology Development Co. | Zhang Y.,Harbin Institute of Technology | And 2 more authors.
International Geoscience and Remote Sensing Symposium (IGARSS) | Year: 2013

Hyperspectral remote sensing is a technique based on the spectroscopy, which contains abundant spectral information besides the spatial information of the images, and overcomes the limitations of the wide-band remote sensing detection. When classifying hyperspectral and multispectral images with the existing algorithms, we use only the spectral information more often. This paper presents an one-class classification techniques, which is based spatial-contextual term, this study modifies the decision function and constraints of support vector data description. Experimental results show that the proposed method achieves good classification performance on hyperspectral image. © 2013 IEEE.


Wang X.,Beijing Twenty First Century Science and Technology Development Co. | Wang X.,Heilongjiang University | Yan Q.,Heilongjiang University | Zhang J.,Harbin Institute of Technology | And 2 more authors.
Zhongguo Jiguang/Chinese Journal of Lasers | Year: 2014

In order to improve the space resolution of hyper-spectral image by fusing the spatial information of multispectral images and the spectral information of hyperspectral images, a hyperspectral image super-resolution algorithm based on relevance vector machine (RVM) is proposed. A brief introduction of the principle of the Price method which fuses multispectral and hyperspectral images to get the super-resolution image is given, and the RVM linear regression is introduced. Combining with the advantages of RVM in regression analysis, a resolution enhancement by revealing the corrspondence of the spatial and spectral information is gotten. The experiment results show that the normalized root-mean-square (RMS) is lower than 0.001 and the spectral angel error is lower than 0.02, which gets a great improvement compared with the results of the Price method and the Elbakary method. The method proposed has a significant result in hyperspectral image reconstruction, which provides a much properer data source for classification, object detection and recognition.


Wang X.,Beijing Twenty First Century Science and Technology Development Co. | Wang X.,Heilongjiang University | Hou C.,Heilongjiang University | Yan Q.,Heilongjiang University | And 3 more authors.
Hongwai yu Jiguang Gongcheng/Infrared and Laser Engineering | Year: 2014

In order to more accurately estimate noise intensity for hyperspectral imagery, the paper proposed a noise estimation algorithm based on relevance vector machine(RVM)for hyperspectral imagery. And the algorithm that used RVM regression, residuals and noise was studied. First of all, this paper introduced the characteristics and shortage of spatial/spectral dimension decorrelation in noise estimation that used widely nowadays for hyperspectral imagery. Then, the nonlinear regression analysis of RVM was introduced. And the residuals will be too large, when there was a strong nonlinear correlation in the system for spatial/spectral dimension decorrelation. To this problem, the paper proposed a new method that used RVM regression to remove strong signal correlation and used the residual images to estimate the noise, so as to improve the stability of the assessment system. Experimental results indicate that the precision of the noise intensity is better than 8%, and show that the method is more effective compared to the traditional method. It concludes that the RVM can satisfy the system requirements of higher precision and stabilization in noise estimation for automatic hyperspectral imagery. ©, 2014, Chinese Society of Astronautics. All right reserved.


Wang X.,Beijing Twenty First Century Science and Technology Development Co. | Wang X.,Heilongjiang University | Zhang J.,Harbin Institute of Technology | Yan Q.,Heilongjiang University | Chi Y.,Beijing Twenty First Century Science and Technology Development Co.
Zhongguo Jiguang/Chinese Journal of Lasers | Year: 2014

Hyperspectral imagery target detection has an important theoretical research value and application prospect, and it is a hot topic in the field of the remote sensing information processing. At present, most detection algorithms need to set an appropriate decision threshold, which is set by hand or computed by using the objects and background information. In practice, the little prior knowledge of the background often limits the application of many algorithms. To solve this problem, a new pure-pixel target detection algorithm for hyperspectral image is presented, which is based on the support vector data description (SVDD). Then the target detection problem is transformed to one-class classification problem. Firstly, SVDD classifier is trained by selected samples, and then the data are classified into inner-class (the target) and outer-class (the background). Next, the spatial characteristics of the target are used to reduce false alarm rate of the classified image. Finally, the ultimate detection results can be obtained. Experimental results of the hyperspectral data show that compared with the two classical spectral angle mapping and constrained energy minimization methods, the proposed method, which only requires a small number of target training samples, can reach the close results as the two algorithms when the optimal threshold values are selected. When the background samples increase, the method is superior to the mentioned two algorithms.

Loading Beijing Twenty First Century Science and Technology Development Co. collaborators
Loading Beijing Twenty First Century Science and Technology Development Co. collaborators