Semantic segmentation of high spatial resolution images with deep neural networks

Availability of reliable delineation of urban lands is fundamental to applications such as infrastructure management and urban planning. An accurate semantic segmentation approach can assign each pixel of remotely sensed imagery a reliable ground object class. In this paper, we propose an end-to-end deep learning architecture to perform the pixel-level understanding of high spatial resolution remote sensing images. Both local and global contextual information are considered. The local contexts are learned by the deep residual net, and the multi-scale global contexts are extracted by a pyramid pooling module. These contextual features are concatenated to predict labels for each pixel. In addition, multiple additional losses are proposed to enhance our deep learning network to optimize multi-level features from different resolution images simultaneously. Two public datasets, including Vaihingen and Potsdam datasets, are used to assess the performance of the proposed deep neural network. Comparison with the results from the published state-of-the-art algorithms demonstrates the effectiveness of our approach.     

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.     

Expanding global mapping of the foliage clumping index with multi-angular POLDER three measurements: Evaluation and topographic compensation

The clumping index measures the spatial aggregation (clumped, random and regular) of foliage elements. The global mapping of the clumping index with a limited eight-month multi-angular POLDER 1 dataset is expanded by integrating new, complete year-round observations from POLDER 3. We show that terrain-induced shadows can enhance bi-directional reflectance distribution function variation and negatively bias the clumping index (i.e. indicating more vegetation clumping) in rugged terrain. Using a global high-resolution digital elevation model, a topographic compensation function is devised to correct for this terrain effect. The clumping index reductions can reach up to 30% from the topographically non-compensated values, depending on terrain complexity and land cover type. The new global clumping index map is compared with an assembled set of field measurements from 32 different sites, covering four continents and diverse biomes.     

Use of service middleware based on ECHO with CSW for discovery and registry of MODIS data

Nowadays, NASA is producing several terabytes Moderate Resolution Imaging Spectroradiometer (MODIS) data everyday; how to find the data with criteria, such as specific times, locations, and scales using an international standard becomes more and more important. In this paper, a service-oriented architecture for use of the integration Earth Observation System Clearing-HOuse (ECHO) with the Open Geospatial Consortium (OGC) Catalogue Service—Web profile (CSW) is put forward. The architecture consists of three roles: a service requester (the user), a service provider (the ECHO metadata server), and a service broker (the GeoNetwork CSW and MODIS registry service middleware). The core component-MODIS registry service middleware includes three components: metadata fetcher, metadata transformer, and metadata register. The metadata fetcher is used to fetch metadata from ECHO metadata server; the metadata transformer is responsible for transform metadata from one form to another; the metadata register is in charge of registering ISO19139-based metadata to CSW. A prototype system is designed and implemented by using the service middleware technology and a standard interface and protocol. The feasibility and the response time of registry and retrieval of MODIS data are evaluated by means of a realistic LPDAAC_ECS MODIS data center. The implementation of this prototype system and the experiment show that the architecture and method is feasible and effective.     

An enhanced bit-wise parallel algorithm for real-time GPS software receiver

Software-based global positioning system (GPS) receivers perform all the baseband signal processing and the high level functions on a general purpose processor. The heavy computational loads of the signal correlation in baseband processing make it difficult for software receivers to operate in real time. In order to improve the real-time performance, an enhanced bit-wise parallel algorithm has been developed in this study. The enhanced algorithm has been implemented and tested in a 12 channels real-time GPS software receiver. The system consists of a radio frequency front end, a data acquisition board and software that runs on a laptop with a Pentium-M 1.5 GHz processor running the Window^® XP operating system. The data acquisition board packs the 2-bit intermediate frequency samples with a 2-bit in/8-bit out shift register and transfers the packed samples to laptop through a USB port. The software running on the laptop performs all the baseband and navigation processing in real time. The test results show that the enhanced algorithm significantly improves the real-time performance of the software receiver by reducing the computational operations for signal correlation by 50% compared with the existing bit-wise parallel algorithm. Furthermore, the enhanced algorithm also reduces the amount of required memory for storing data for signal correlation.     

GPS-RTK与全站仪结合在丘陵地区公路断面测量中的应用

介绍了GPS-RTK和全站仪测量技术的特点及二者结合测量公路断面的方法,通过实例论述了该方法运用在丘陵地区公路测量中的实用性和高效性,并进行了精度分析。     

一定分布模式下的最优Lp估计

观测数据在一定的分布模式下,需从理论上解决究竟用多大的ｐ值进行Ｌｐ估计（最优ｐ值的确定）,本文基于渐近方差整体最小的原则,给出了最优Ｌｐ估计的定义。观测数据中含有粗差（异常值）的扰动,其误差分布可视为污染分布,作者分析了最优ｐ值的大小,结果表明,若观测数据中含有不同大小、数量的粗差,为使Ｌｐ估计差最小,则ｐ将取［１．２,１．５］中的某一定值,Ｌｐ估计结果最优,此结论对测量数据处理具有一定的参考价值     

GPS技术在地球科学中的应用和展望

GPS(全球定位系统)能为全球提供高精度、全天候、连续、实时的三维定位、三维测速和时间基准。本文对它在地球科学中的应用和可能的发展作一扼要介绍。其中,包括了它在几何大地测量、物理大地测量、地球动力学、海洋大地测量、测时和授时、工程测量、导航等方面的应用及其展望。     

粗差数学期望平移模型的理论研究

现有文献对数学期望平移模型的理论分析仅考虑了观测值统计独立的特殊情况。基于观测值统计相关的一般情况,导出了数学期望平移参数估值(▽S_i)的简明表达式。在此基础上,采用统计预测理论对▽S_i进行了直观的理论解释,扩展了统计学文献中的有关结论。借助于实例,分析了▽S_i与最小二乘残差的本质区别。     

地球空间信息技术在新一轮国土资源大调查中的应用探讨

从分析全球变化和社会可持续发展研究的要求出发,提出地球空间信息技术的形成背景和应用基础。针对新一轮国土资源大调查的性质和任务,探讨了地球空间信息技术在新一轮国土资源大调查中的应用。