SAR image despeckling with a multilayer perceptron neural network

Speckle noise in synthetic-aperture radar (SAR) images severely hinders remote sensing applications; therefore, the appropriate removal of speckle noise is crucial. This paper elaborates on the multilayer perceptron (MLP) neural-network model for SAR image despeckling by using a time series of SAR images. Unlike other filtering methods that use only a single radar intensity image to derive their parameters and filter that single image, this method can be trained using archived images over an area of interest to self-learn the intensity characteristics of image patches and then adaptively determine the weights and thresholds by using a neural network for image despeckling. Several hidden layers are designed for feedforward network training, and back-propagation stochastic gradient descent is adopted to reduce the error between the target output and neural-network output. The parameters in the network are automatically updated in the training process. The greatest advantage of MLP is that once the despeckling parameters are determined, they can be used to process not only new images in the same area but also images in completely different locations. Tests with images from TerraSAR-X in selected areas indicated that MLP shows satisfactory performance with respect to noise reduction and edge preservation. The overall image quality obtained using MLP was markedly higher than that obtained using numerous other filters. In comparison with other recently developed filters, this method yields a slightly higher image quality, and it demonstrates the powerful capabilities of computer learning using SAR images, which indicate the promising prospect of applying MLP to SAR image despeckling.     

Building a model-based personalised recommendation approach for tourist attractions from geotagged social media data

When travelling, people are accustomed to taking and uploading photos on social media websites, which has led to the accumulation of huge numbers of geotagged photos. Combined with multisource information (e.g. weather, transportation, or textual information), these geotagged photos could help us in constructing user preference profiles at a high level of detail. Therefore, using these geotagged photos, we built a personalised recommendation system to provide attraction recommendations that match a user's preferences. Specifically, we retrieved a geotagged photo collection from the public API for Flickr (Flickr.com) and fetched a large amount of other contextual information to rebuild a user's travel history. We then created a model-based recommendation method with a two-stage architecture that consists of candidate generation (the matching process) and candidate ranking. In the matching process, we used a support vector machine model that was modified for multiclass classification to generate the candidate list. In addition, we used a gradient boosting regression tree to score each candidate and rerank the list. Finally, we evaluated our recommendation results with respect to accuracy and ranking ability. Compared with widely used memory-based methods, our proposed method performs significantly better in the cold-start situation and when mining ‘long-tail’ data.     

Recent insights into physiological responses to nutrients by the cylindrospermopsin producing cyanobacterium,Cylindrospermopsis raciborskii

The harmful cyanobacterium Cylindrospermopsis raciborskii is a widespread species increasingly being recorded in freshwater systems around the world. Studies have demonstrated some key attributes of this species which may explain its global dominance. It has a high level of flexibility with respect to light and nutrients, being capable of growth under low and variable light conditions. However, it is the strategy with respect to nutrient utilization that has received more attention. Unlike many bloom forming species, the dominance of this species is not simply linked to higher nutrient loads. In fact it appears that it is more competitive when phosphorus and nitrogen availability is low and/or variable. An important component of this flexibility appears to be the result of within-population strain variability in responses to nutrients, as well as key physiological adaptations. Strain variability also appears to have an effect on the population-level cell quota of toxins, specifically cylindrospermopsins(CYNs). Field studies in Australia showed that populations had the highest proportion of toxic strains when dissolved inorganic phosphorus was added, resulting in stoichiometrically balanced nitrogen and phosphorus within the cells. These strategies are part of an arsenal of responses to environmental conditions, making it a challenging species to manage. However, our ability to improve bloom prediction will rely on a more detailed understanding of the complex physiology and ecology of this species.     

基于GPS速度场的云南地壳应力积累特征分析

利用2009~2015年的GPS水平运动速度场数据,解算云南地区的地壳应变场,在红河断裂与曲江断裂选取两个GPS剖面,计算并分析两个断裂的应变积累特征。结果显示：1）红河、曲江、小江等多个主要断裂应变场存在张、压交替的时空演化特征,近期曲江断裂南段表现为东西向拉张,红河断裂北段东西向拉张量值较大,约12.0×10-8/a,云南南部主要表现为NNE向的压性变化;2）从GPS剖面看,红河、曲江断裂在两个方向上均显示右旋走滑与拉张的变化特征,从量值上看,红河断裂北段运动速率较大,约8.90 mm/a。     

一种自然邻近关系查询的空间索引结构

自然邻近关系在空间数据库管理系统中具有十分重要的地位。为了克服Voronoi图受空间目标位置的影响,本文采用约束Delaunay三角网,针对在二维欧式空间中离散面目标的自然邻近关系查询问题。首先对离散面进行三角剖分,然后将代表相同邻近关系的三角形合并成一个union,进而建立面目标自然邻近关系表达结构-Unions Delaunay。通过与Unions Delaunay相结合建立空间索引结构-R-tree Gridfile来实现对空间目标自然邻居完备的候选集提取以及索引结构的更新。对比实验结果表明R-tree Gridfile索引结构能够有效支持空间目标的自然邻近关系查询。     

Assessment of prospective hazards resulting from the 2017 earthquake at the world heritage site Jiuzhaigou Valley,Sichuan, China

On August 8, 2017, a Ms = 7.0 magnitude earthquake occurred in the Jiuzhaigou Valley, in Sichuan Province, China(N: 33.20°, E: 103.82°). Jiuzhaigou Valley is an area recognized and listed as a world heritage site by UNESCO in 1992. Data analysis and field survey were conducted on the landslide, collapse, and debris flow gully, to assess the coseismic geological hazards generated by the earthquake using an unmanned aerial vehicle(UAV), remote-sensing imaging, laser range finders, geological radars, and cameras. The results highlighted the occurrence of 13 landslides, 70 collapses, and 25 potential debris flow gullies following the earthquake. The hazards were classified on the basis of their size and the potential property loss attributable to them. Consequently, 14 large-scale hazards, 30 medium-sized hazards, and 64 small hazards accounting for 13%, 28%, and 59% of the total hazards, respectively, were identified. Based on the variation tendency of the geological hazards that ensued in areas affected by the Kanto earthquake(Japan), Chi-chi earthquake(Taiwan China), and Wenchuan earthquake(Sichuan China), the study predicts that, depending on the rain intensity cycle, the duration of geological hazard activities in the Jiuzhaigou Valley may last over ten years and will gradually decrease for the following five to ten yearsbefore returning to pre-earthquake levels. Thus,necessary monitoring and early warning systems must be implemented to ensure the safety of residents,workers and tourists during the construction of engineering projects and reopening of scenic sites to the public.     

Evaluating the suitability of different terrains for sustaining human settlements according to the local elevation range in China using the ASTER GDEM

The topographical suitability assessment of human settlements(SAHS) creates a solid foundation for regional population distribution and socio-economic development. Local elevation range(LER) is an important factor that can be used to assess the suitability of different terrains for sustaining human settlements. However, current digital elevation model(DEM)-based LER products suffer from some challenges typically because of their subjectively selected neighborhood scales and coarser spatial resolution. In this study, we initially determined the optimal statistical window and then calculated the appropriate LER with the finer resolution data of the Advanced Spaceborne Thermal Emission and Reflection Radiometer Global DEM(ASTER GDEM) products for China. Then, the appropriate LER was used to evaluate the topographical SAHS and its correlations with the national gridded population distribution(1 km × 1 km) in 2010. The results show that the optimal statistical window for calculating a 1 arc-second(about 30 m) resolution GDEM LER for China can be determined using a 51 × 51 grid unit(width × height) within a rectangular neighborhood, corresponding to an area of about 2.34 km~2. Secondly, the LER values in the southern and western China were greater than those of the north and east, showing a trend which consistently reflects the general spatial features of landforms. Finally, the relationship between GDEM LER and population density was highly correlated with the R~2 value of 0.81. It showed that 85.22% of the Chinese population was located in areas where the LER is lower than 500 m. The topographically suitable area within China decreased from the southeastern coastal zone towards the northwestern inland areas due to transition from plains and basins to plateaus and mountains. The total area of moderate to high suitable level was 423.84 × 10~4 km~2, or 44.15% of the total land area, with 88.17% of the national population. Our study demonstrates the usefulness of appropriate LER in evaluating the topographical SAHS as well as its significant impact on population distribution.     

等效沙厚度研究进展

风、沉积物和地表覆盖是控制沙丘地貌的三大要素.沉积物尤其是沙源供应对沙丘地貌的影响仍有待深入研究.沙丘体积的研究促使了等效沙厚度(equivalent sand thickness,EST)概念的提出.EST研究经历了多个沙丘和亚沙丘尺度2个研究阶段.多个沙丘EST研究发现,EST与沙丘高度之间具有较好的一致性,其与方...     

贵州矿山地质环境恢复治理效果评估方法改进研究——以贵州省毕节市某煤矿为例

在收集矿山地质环境恢复治理相关规范和技术要求的基础上,结合矿山开采造成的地质环境问题的现状、分布,从众多常用的矿山地质环境恢复治理评价指标中,选取普适性较强的地质灾害治理效果、水环境恢复效果、景观植被恢复效果、管护措施这4个评价指标作为评价因子,从而验证上述普适性较强的评价因子可以较好的反映治理恢复效果。根据评价因子的重要性,分别赋予不同的权重,采用加权综合法进行计算,对恢复治理效果评估方法进行改进量化,综合判定地质环境恢复治理效果。本文以贵州省毕节市某煤矿为例,采用的矿山地质环境恢复治理评估方法通过验证,可以为同类型矿山开展地质环境恢复治理效果评估提供借鉴。     

中国锂矿的多旋回深循环内外生一体化成矿理论及其找矿应用

中国锂资源的分布具有卤水型与硬岩型相伴相随、若即若离的特点。“多旋回深循环内外生一体化”成锂理论,在多年三稀金属矿产找矿实践和理论研究的基础上不断丰富完善。中国锂矿的形成与“多旋回”构造运动密切相关,从前寒武纪到新生代均有成矿潜力,可构成一个完整的多旋回成矿谱系。锂的“深循环”,一是锂深度参与成岩成矿的物质循环;二是需要一个“圈闭”的构造背景将锂“捕获”以避免其过度分散,锂从开始加入到岩浆与最终定位的深度之差,是硬岩型锂矿成矿的关键之一,压差越大越有利于伟晶岩型锂矿的形成。大量锂矿实例显示锂的物质循环是“内外生一体化”的统一过程,高海拔地区(山上)的含锂地质体(花岗岩类甚至直接就是含锂矿床)经风化剥蚀之后,可能成为沉积型锂矿的物质来源之一;而富含锂的沉积岩经过埋藏、变质、深熔也可以形成含锂的岩浆岩、伟晶岩。我国西部塔里木盆地、四川盆地、扎布耶盆地及东部的江汉盆地、吉泰盆地、周田盆地等大小不一的盆地均含锂,而其周边造山带中也不同程度发育硬岩型锂矿,这就为区域找矿指明了方向。“多旋回深循环内外生一体化”成锂理论是三稀矿产成矿理论的重要组成部分,为我国锂矿找矿工作提供了指导和借鉴,在甲基卡、可尔因、阿尔金、幕阜山等锂矿矿集区的找矿实践中发挥了积极作用。