基于RS、GIS的东喜马拉雅南迦巴瓦峰地区断裂构造的定量研究

东喜马拉雅南迦巴瓦峰地区地质构造十分复杂,对该区板块缝合带的空间位置问题,长期以来,一直认为沿雅鲁藏布江呈弧形展布。笔者以ETM+为主要信息源,以遥感与GIS为技术支撑,对南峰地区的断裂构造进行了详细的遥感解译分析,从构造统计分析的角度对断裂构造进行定量研究,从分维值及趋势值的空间分布及变化特征分析,认为缝合带的南东段、北段大致沿雅鲁藏布江展布,而北西段位于雅鲁藏布江的北西侧,沿东久-米林(断裂带)展布,这也与近几年来开展了1∶25万区域地质调查获得的新认识相一致。     

基于图像的天体搜索研究与自动化集成软件开发

天体搜索是天文数据处理流程的一个重要环节,也是以平方公里阵列射电望远镜(Square Kilometre Array, SKA)为代表的下一代射电望远镜在面向海量数据处理中的挑战之一.现今天体自动搜索算法、软件已日趋成熟并投入应用,不过在自动化、兼容性等方面仍具有提升空间.以更自动化、更适应海量数据需求的天体搜索算法研究为宗旨,以现有算法为研究基础,天体自动搜索软件系统得到设计和开发.该系统包含友好的交互式用户操作界面,具备可视化输出数据显示、兼容不同数据输入和输出并包含为实际应用服务的文件管理功能.该系统对于大天区图以及图像集,均能够很好地进行自动化处理.测试结果显示,上述方法对于天体搜索的改进有一定成效.后续将在此基础上对该集成系统做进一步的改进开发,以适应更多的需求.     

Research on Parallel Algorithms for uv-faceting Imaging

The uv-faceting imaging is one of the widely used large field of view imaging technologies, and will be adopted for the data processing of the low-frequency array in the first stage of the Square Kilometre Array (SKA1). Due to the scale of the raw data of SKA1 is unprecedentedly large, the efficiency of data processing directly using the original uv-faceting imaging will be very low. Therefore, a uv-faceting imaging algorithm based on the MPI (Message Passing Interface)+OpenMP (Open Multi-Processing) and a uv-faceting imaging algorithm based on the MPI+CUDA (Compute Unified Device Architecture) are proposed. The most time-consuming data reading and gridding in the algorithm are optimized in parallel. The verification results show that the results of the proposed two algorithms are basically consistent with that obtained by the current mainstream data processing software CASA (Common Astronomy Software Applications), which indicates that the proposed two algorithms are basically correct. Further analysis of the accuracy and total running time shows that the MPI+CUDA method is better than the MPI+OpenMP method in both the correctness rate and running speed. The performance test results show that the proposed algorithms are effective and have certain extensibility.     

Detection of spatio-temporal evolutions on multi-annual satellite image time series: A clustering based approach

The expansion of satellite technologies makes remote sensing data abundantly available. While the access to such data is no longer an issue, the analysis of this kind of data is still challenging and time consuming. In this paper, we present an object-oriented methodology designed to handle multi-annual Satellite Image Time Series (SITS). This method has the objective to automatically analyse a SITS to depict and characterize the dynamic of the areas (the way that the land cover of the areas evolve over time). First, it identifies the spatio-temporal entities (reference objects) to be tracked. Second, the evolution of such entities is described by means of a graph structure and finally it groups together spatio-temporal entities that evolve similarly. The analysis were performed on three study areas to highlight inter (among the study areas) and intra (inside a study area) similarity by following the evolution of the underlying phenomena. The analysis demonstrate the benefits of our methodology. Moreover, we also stress how an expert can exploit the extracted knowledge to pinpoint relevant landscape evolutions in the multi-annual time series and how to make connections among different study areas.     

SAR影像中叠掩与阴影区域的识别 ——以湖北巴东为例

SAR影像中叠掩和阴影区域的识别是滑坡监测中的首要工作,本文针对这一工作利用雷达卫星成像时的几何模型与图像处理中形态学方法综合识别叠掩与阴影区域,以ALOS-2数据为例,对比该地区雷达强度图像,验证了本文试验方法的可靠性。     

结合深度学习和图割法的遥感影像建筑物检测

提出一种基于卷积神经网络和图割法的自动提取高分影像建筑物的方法。首先,通过卷积神经网络定位与检测建筑物的位置,逐一提取单个建筑物轮廓,利用检测结果分别建立建筑物和非建筑物的高斯混合模型（GMM）,然后结合最大流最小割的图像分割方式实现全局优化,完成建筑物初步提取,最后用形态学进行优化。通过试验证明了该方法的可行性。     

一种高光谱遥感图像快速谱聚类算法

为获得分类效果更优良的遥感图像分类方式并解决高光谱遥感图像分类运算速度缓慢的问题,集成Lanczos算法与谱聚类算法,探讨了高光谱遥感图像谱聚类算法应用于遥感图像分类的可行性,提出了一种面向高光谱遥感图像的快速谱聚类算法;通过对比美国圣地亚哥机场高光谱遥感图像K-均值算法与谱聚类算法的分类结果,发现面向高光谱遥感图像的谱聚类算法易于识别线性地物,且分类的速度能得到较大提升。     

深度卷积神经网络支持下的遥感影像飞机检测

针对YOLOv3算法对小目标检测较差及出现较多漏检的问题,本文提出了一种优化的YOLOv3算法.首先使用K-means算法计算出与数据集相适用的锚框;其次将扩张卷积引入到YOLOv3网络,用来增强网络高层的感受野,改善小目标的检测效果;然后使用深度可分离卷积取代YOLOv3网络残差模块中的普通卷积,可减少计算量,从而得到一种新型卷积神经网络结构;最后在数据集上进行对比试验.结果表明,优化的YOLOv3算法能够检测出更多目标,降低漏检率,相比于YOLOv3算法,其召回率提高11.86%,F1-score提高2.99%.     

顾及局部与结构特征的稀疏多项式逻辑回归高光谱图像分类方法

稀疏多项式逻辑回归在分类中仅利用图像光谱信息,导致分类效果不太理想.本文提出了一种顾及局部与结构特征的稀疏多项式逻辑回归高光谱图像分类方法.首先利用加权均值滤波与拓展形态学多属性剖面对原始高光谱图像进行局部与结构特征提取;然后对二者进行加权平均特征级融合以获取更具唯一性的像元特征;最后由稀疏多项式逻辑回归分类器对融合结果进行分类.结果表明,本文方法能有效地提高分类精度,而且具有较强的稳健性.     

改进分数阶达尔文粒子群的多Renyi熵图像分割算法

针对智能优化图像分割算法易陷入局部最优、分割精度不高等问题,本文融合改进的分数阶达尔文粒子群算法和二维Renyi熵多阈值,提出了一种新的多阈值遥感图像分割算法。算法利用粒子自身进化信息来定义进化因子,结合进化因子并利用高斯图函数调整分数阶次a系数以实现精确计算和快速收敛;根据局部最优概率因子对局部最优位置进行Levy飞行随机扰动以提高算法跳出局部最优的能力;同时将二维Renyi熵单阈值扩展到多阈值分割上,并结合改进的分数阶达尔文粒子群算法,将二维Renyi熵多阈值应用于遥感图像分割中仿真结果表明,与其他2种智能优化分割算法相比,本文分割算法在细节处理和分割精度上均有明显优势,在PRI上至少提升7.27%、VOI至少降低6.5%、GCE至少降低10.4%.