三维地形可视化中不同图元渲染效率比较分析

大规模地形场景一般由几百万甚至更多的顶点构成,地形瓦片的图元组织与绘制效率直接决定了地形渲染的速度和质量。为明确Triangles,Triangle Strip,Triangle Fan和优化的Triangle Strip图元渲染效率,选取常用的11种大小瓦片样本进行对比试验。从初始化时间、索引缓存大小和实时绘制时间3个方面评估图元效率。结果表明：在地形瓦片小于256时,Triangles和优化的Triangle Strip具备较高的渲染效率,Triangles会占用较大的索引缓存;在地形瓦片超过256时,优化后的Triangle Strip图元具有最高绘制效率,并且占用最小的索引缓存。     

典型遥感数据分类方法的适用性分析——以遥感图像场景分类为例

目前普遍采用的分类器通常都是针对单一或小量任务而设计的,在小数据量的处理中能取得比较满意的结果。但对于海量遥感数据的处理,其在处理时效和分类精度方面还有待研究。本文以遥感图像场景分类任务为例,着重对遥感数据分类问题中几种典型分类方法的适用性进行比较研究,包括K近邻(KNN)、随机森林(RF),支持向量机(SVM)和稀疏表达分类器(SRC)等。分别从参数敏感性,训练样本数据量,待分类样本数据量和样本特征维数对分类器性能的影响等几个方面进行比较分析。实验结果表明:(1)KNN,RF和L0-SRC方法相比RBF-SVM,Linear-SVM和L1-SRC,受参数影响的程度更弱;(2)待分类样本固定的情况下,随着训练样本数目的增加,SRC类型分类方法的分类性能最佳,SVM类型方法次之,然后是RF和KNN,在总体分类时间上呈现出L0-SRCL1-SRCRFRBF-SVM/Linear-SVMKNN/L0-SRC-Batch的趋势;(3)训练样本固定的情况下,所有分类方法的分类精度几乎都不受待分类样本数目变化的影响,RBF-SVM方法性能最佳,其次是L1-SRC,然后是Linear-SVM,最后是RF和L0-SRC/L0-SRC-Batch,在总体分类时间上,L1-SRC和L0-SRC相比其他分类方法最为耗时;(4)样本特征维数的变化不仅影响分类器的运行效率,同时也影响其分类精度,其中SRC和KNN分类器器无需较高的特征维数即可获得较好的分类结果,SVM对高维特征具有较强的包容性和学习能力,RF分类器对特征维数增加则表现得并不敏感,特征维数的增加并不能对其分类精度的提升带来更多的贡献。总的来说,在大数据量的遥感数据分类任务中,现有分类方法具有良好的适用性,但是对于分类器的选择应当基于各自的特点和优势,结合实际应用的特点进行权衡和选择,选择参数敏感性较小,分类总体时间消耗低但分类精度相对较高的分类方法。     

《遥感学报》20年:从热点到前沿

为反映《遥感学报》及中国遥感学科近20年的发展趋势,针对《遥感学报》在1997年—2015年期间出版的所有1804篇论文进行了统计分析,给出了国内高校及研究机构的发文量排名和受国家自然科学基金资助的发文情况,并对比了常设栏目和专题栏目的论文发文量及被引用情况。通过比较4个时间段内论文关键词的共现关系,反映出随着新的卫星载荷不断发射,观测技术由单一观测变为多源卫星观测,遥感定量反演模型由简单变得更为复杂,遥感技术应用也由最初的测绘、国土调查等单一应用逐渐变为多学科交叉应用,特别是在近年来国内外社会热点的驱动下,遥感技术已开始在灾害应急、全球变化、大气污染、粮食安全等领域发挥出越来越重要的作用。此外,无论是在当前遥感学科发展还是国家需求的现状下,都迫切需要建立国家级的遥感应用综合信息系统,以提高对环境和资源的宏观调控能力,为中国经济和社会可持续发展战略、布局和趋势预测,为资源管理、环境保护、防灾减灾以及实现资源环境、经济、社会的宏观调控,提供科学的数据和决策支持。     

地表覆盖遥感产品更新完善的研究动向

近年来,多尺度地表覆盖遥感产品的不断涌现,为环境变化研究、地球系统模拟、地理国(世)情监测和可持续发展规划等提供了重要科学数据。为更好地满足广大用户日益增长的应用需求,应对地表覆盖遥感产品进行持续更新完善,保持其时效性、增强时序性、丰富多样性。针对大面积地表覆盖遥感产品更新完善所面临的主要问题,介绍和评述了国内外有关研究动向,包括影像与众源信息相结合的更新、数据类型细化与完善、地表覆盖真实性验证,并作了简要展望。     

热带森林植被的动态变化遥感监测

本文讨论了以热带森林植被为主体的再生资源的面积动态变化监测。研究中包括两个部分。首先,我们利用多时相遥感图像对大面积的西双版纳州进行地类判读,系统地分析了森林植被的动态变化。其次,利用Landsat MSS和TM数据对自然保护区的动态变化进行了包含无监督分类和归一化差值植被指数分析的数字图像处理,变化分类也相当符合实际。总的实验结果表明,这种监测方法是很有效的,可在再生资源监测中特别是在森林植被监测中加以推广应用。     

Assessing soil salinity using WorldView-2 multispectral images in Timpaki,Crete, Greece

Salinization is one of the major soil degradation threats occurring worldwide. This study evaluates the feasibility of operational surface soil salinity mapping based on state-of-the-art Earth Observation (EO) products captured by sensors on-board WorldView-2 (WV2) and Landsat 8 satellites. The proposed methods are tested in Timpaki, south-central Crete,Greece, where brackish water irrigation puts soil health at risk of soil salinization. In all cases, EO products are calibrated against soil samples collected from bare soil locations. Results indicate a moderate correlation of observed EC_e values with the investigated remote sensing parameters. Regarding sensitivity to saline soil, the yellow band displays higher values. Comparison between methods used in the literature shows that those developed specifically for soil salinity, and especially index S5, perform better. The proposed ‘detection index’ and 3D PCA transformation methodology perform reasonably well in detecting areas with high EC_e values and provide a simple and effective operational alternative for saline topsoil detection and mapping.     

Use of LiDAR-derived DEM and a stream length-gradient index approach to investigation of landslides in Zagros Mountains,Iran

This paper presents an approach to stream length-gradient index analysis to identify tectonic signatures. The graded profile of the Dez River in Zagros Mountains, Iran, indicates that the area has been tectonically disturbed, and it triggers landslide hazards. The high-gradient index shows that a steeper gradient could be potentially a signature for landslides identification. The digital surface models acquired by airborne LiDAR were used in this study to generate the HRDEM. Our result shows a great potential for improving landslide investigations by implementing stream length-gradient index derived from the HRDEM in conjunction with the landslide inventories data-set in the GIS environment. We also identified a correlation between the stream length-gradient index and the graded topographic profile with slopes and landslides. This empirical approach was verified by geodata analytics and landslide inventories data-set in conjunction with field observations. This study has identified the locations of high-gradient indices with susceptible to landslides.     

Region of interest extraction via common salient feature analysis and feedback reinforcement strategy for remote sensing images

Saliency detection is an effective approach to extract regions of interest (ROIs) for remote sensing images. However, existing saliency detection models mainly focus on ROI extraction from a single image and usually are not able to produce satisfactory results because of complex background interference in remote sensing images. The employment of mutual information in a set of remote sensing images can provide an effective solution to this issue. In this paper, we propose a novel saliency detection model for multiple remote sensing images to simultaneously extract ROIs and identify images without ROIs. First, common salient feature analysis based on synthesized feature clustering and global contrast is implemented to exploit global correspondence in the synthesis feature domain, thereby highlighting preliminary ROIs against background interference and assigning lower saliency values to images without ROIs. Then, we design an exclusion criterion based on saliency value judgment to remove images without ROIs, and the remaining saliency maps are refined by an enhancement strategy. Finally, the enhanced maps serve as a feedback to yield a homogenous synthesized feature space in which integral ROIs with subtle borders are extracted by the reused cluster-based saliency calculation. Experiments reveal that our model outperforms seven state-of-the-art models by achieving the best ROC curve (AUC = 0.945) and maximal F-measure at 0.729.     

Effect of flying altitude and pulse repetition frequency on laser scanner penetration rate for digital elevation model generation in a tropical forest

In tropical forests, the penetration ability of airborne laser scanning (ALS) may be limited because of highly dense vegetation cover. However, in the typical planning of ALS surveys, the ability of laser pulses to penetrate forests is not considered. Nine round-trip flight lines covering the area of a tropical forest on the northeast side of the Tsengwen Reservoir in Taiwan were designed in this study. Five flight lines flew at altitudes of 1.525, 1.830, 2.135, 2.440, and 2.745 km, and the other four had pulse repetition frequencies (PRFs) of 100, 150, 200, and 250 kHz. The laser penetration index (LPI) is a quantitative index measuring the penetration ability of the ALS and consists of the ratio of the number of laser pulses reaching the forest floor to the total number of laser pulses. The LPI was used to represent the laser penetration rate and investigate the influence of flying altitude and PRF on the LPI. The results showed that as the flying altitude decreased by 1 km, the average LPI increased by 10%, and as the PRF decreased by 50 kHz, the average LPI increased by 2%. The effect of the LPI on digital elevation models (DEMs) was confirmed by visual images obtained by DEMs at five altitudes. The DEM obtained at an altitude of 2.745 km was coarsely textured, whereas that obtained at an altitude of 1.525 km was finely textured. The in-situ height data obtained from the electronic Global Navigation Satellite System (eGNSS) were compared with the data of the ALS-generated DEMs. The results indicated that when the LPI ≥60%, the height difference between the in situ data and DEM data was not prominent. However, when the LPI <60%, the ALS-derived DEM could be higher or lower than the in-situ height; the largest difference between the two was 1.7 m. The LPI of a forest should be considered for ALS survey planning, especially when consistent DEM precision for large tropical forest areas is paramount.     

Mapping fractional landscape soils and vegetation components from Hyperion satellite imagery using an unsupervised machine-learning workflow

An unsupervised machine-learning workflow is proposed for estimating fractional landscape soils and vegetation components from remotely sensed hyperspectral imagery. The workflow is applied to EO-1 Hyperion satellite imagery collected near Ibirací, Minas Gerais, Brazil. The proposed workflow includes subset feature selection, learning, and estimation algorithms. Network training with landscape feature class realizations provide a hypersurface from which to estimate mixtures of soil (e.g. 0.5 exceedance for pixels: 75% clay-rich Nitisols, 15% iron-rich Latosols, and 1% quartz-rich Arenosols) and vegetation (e.g. 0.5 exceedance for pixels: 4% Aspen-like trees, 7% Blackberry-like trees, 0% live grass, and 2% dead grass). The process correctly maps forests and iron-rich Latosols as being coincident with existing drainages, and correctly classifies the clay-rich Nitisols and grasses on the intervening hills. These classifications are independently corroborated visually (Google Earth) and quantitatively (random soil samples and crossplots of field spectra). Some mapping challenges are the underestimation of forest fractions and overestimation of soil fractions where steep valley shadows exist, and the under representation of classified grass in some dry areas of the Hyperion image. These preliminary results provide impetus for future hyperspectral studies involving airborne and satellite sensors with higher signal-to-noise and smaller footprints.