Terrain synthesis from digital elevation models

In this paper, we present an example-based system for terrain synthesis. In our approach, patches from a sample terrain (represented by a height field) are used to generate a new terrain. The synthesis is guided by a user-sketched feature map that specifies where terrain features occur in the resulting synthetic terrain. Our system emphasizes large-scale curvilinear features (ridges and valleys) because such features are the dominant visual elements in most terrains. Both the example height field and user's sketch map are analyzed using a technique from the field of geomorphology. The system finds patches from the example data that match the features found in the user's sketch. Patches are joined together using graph cuts and Poisson editing. The order in which patches are placed in the synthesized terrain is determined by breadth-first traversal of a feature tree and this generates improved results over standard raster-scan placement orders. Our technique supports user-controlled terrain synthesis in a wide variety of styles, based upon the visual richness of real-world terrain data.     

Advances in the data compression of digital elevation models

The maintenance and dissemination of spatial databases requires efficient strategies for handling the large volumes of data that are now publicly available. In particular, satellite and aerial imagery, radar, LiDAR, and digital elevation models (DEMs) are being utilised by a sizeable user-base, for predominantly environmental applications. The efficient dissemination of such datasets has become a key issue in the development of web-based and distributed computing environments. However, the physical size of these datasets is a major bottleneck in their storage and transmission. The problem is often exaggerated when the data is supplied in less efficient, proprietary or national data formats.This paper presents a methodology for the lossless compression of DEMs, based on the statistical correlation of terrain data in local neighbourhoods. Most data and image compression algorithms fail to capitalise fully on the inherent redundancy in spatial data. At the same time, users often prefer a uniform solution to all their data compression requirements, but these solutions may be far from optimal. The approach presented here can be thought of as a simple pre-processing of the elevation data before the use of traditional data compression software frequently applied to spatial data sets, such as GZIP. Identification and removal of the spatial redundancy in terrain data, with the use of optimal predictors for DEMs and optimal statistical encoders such as Arithmetic Coding, gives even higher compression ratios. Both GZIP and our earlier approach of combining a simple linear prediction algorithm with Huffman Coding are shown to be far from optimal in identifying and removing the spatial redundancy in DEMs. The new approaches presented here typically halve the file sizes of our earlier approach, and give a 40–62% improvement on GZIP-compressed DEMs.     

Accuracy assessment of automatically derived digital elevation models from aster data in mountainous terrain

The Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) aboard the Terra satellite was designed to generate along‐track stereo images. The data are available at low cost, providing a feasible opportunity for generating digital elevation models (DEMs) in areas where little or no elevation data are yet available. This study evaluates the accuracy of DEMs extracted from ASTER data covering mountainous terrain. For an assessment of the achieved accuracies in the Andean study site, comparisons were made to similar topographical conditions in Switzerland, where reference data were available. All raw DEMs were filtered and interpolated by the post‐processing tools included with PCI Geomatica, the software package used. After carefully checking the DEM quality, further post‐processing was undertaken to eliminate obvious artefacts such as peaks and sinks. Accuracy was tested by comparing the DEMs in the Swiss Alps to three reference models. The achieved results of the generated DEMs are promising, considering the extreme terrain. Given accurate and well‐distributed ground control points (GCPs), it is possible to generate DEMs with a root mean square (RMS) error between 15?m and 20?m in hilly terrain and about 30?m in mountainous terrain. The DEMs are very accurate in nearly flat regions and on smooth slopes with southern expositions: errors are generally within ±10?m in those cases. Larger errors do appear in forested, snow covered or shady areas and at steep cliffs and deep valleys with extreme errors of a few hundred metres. The evaluation showed that the quality of the DEMs is sufficient for enabling atmospheric, topographic and geometric correction to various satellite datasets and for deriving additional products.     

Geomorphometric processing of digital elevation models

The general system of geomorphometry is composed of elevation, derivatives of elevation at a point, and moments of the distribution of elevation over some area. All of the point measures in this system can be obtained by computer processing of a digital elevation model (DEM), and they can be used as input to the analysis and classification of terrain. A suite of FORTRAN programs implementing this system for dense grid DEMs has been designed and used in various operating environments. Attention has been given to the methods used in the approximation of terrain concepts such as slope and relief. An area of high relief in subarctic Canada is used to illustrate the discussion.     

Errors in river lengths derived from raster digital elevation models

Length of river reaches is one of the most important characteristics of stream networks when applying hydrological or environmental simulation models. A common method of obtaining estimates of river lengths is based on deriving flow directions, accumulated area and drainage lines from raster digital elevation models (DEM). This method leads to length estimates with variable accuracy, which depends on DEM horizontal resolution, flatness of terrain, DEM vertical accuracy, the algorithm used to obtain flow directions and the way by which distances are calculated over raster structures. We applied an automatic river length extraction method for eight river reaches in the River Uruguay Basin (206 000 km²), in Southern Brazil, and compared its results to the lengths obtained from drainage vector lines digitalized over satellite images. Our results show that relative errors can be higher than 30% in flat regions with relatively low DEM resolution. Preprocessing of DEM by the method known as stream burning greatly improves results, reducing errors to the range 1.9–7.4%. Further improved estimates were obtained by applying optimized values for the length of orthogonal and diagonal steps called distance transforms, reducing the errors to the range ?2.0–3.3%.     

Extraction of bajadas from digital elevation models and satellite imagery

A methodology was designed for the extraction of bajadas from the 15 min US Geological Survey digital elevation models and Landsat Thematic Mapper imagery. The method was demonstrated for the Death Valley-California where progressive eastward tilting has enabled the west-side fans to coalesce and form bajadas. First, the drainage that crossed the uplands and the bajadas was extracted from the DEM. The drainage pixels were successively grown by checking the surrounding pixels on the basis of their gradient. It was concluded that for gradient in the interval [2°,11°] the upslope bajadas border was segmented. In order to eliminate the drainage pixels that belonged to the uplands, the drainage pixels were subtracted. Then, the isolated small 8-connected foreground pixels were identified and subtracted too. Finally, region growing was performed again to the remaining pixels with the same growing criterion. Isolated 8-connected background pixels, representing almost flat regions inside bajadas, were identified and merged to the segmented pixels. At the end, by taking into account the spectral response in the satellite image, the downslope border of bajadas was segmented. The extracted polygon was in agreement with the information depicted on (a) the US Geological Survey topographic map of scale 1:100,000 and (b) the satellite image and (c) the polygon classified manually by a photo-interpreter.     

Morphological operators to extract channel networks from digital elevation models

The automated extraction of drainage networks includes generation and processing of digital elevation models (DEMs) obtained from the remotely sensed data having stereo viewing capability. The latter aspect generally aims to extract terrain features such as elevation contours and channel networks. In this technical note, the application of morphological operators to extract channel networks from the digital elevation model is described. The methodology is illustrated using a transcendentally generated DEM that bears the spatially distributed regions in grey levels, assumed as the regions of topographic reliefs and the V-shaped crenulations in successive elevation contours. The authors conclude that the adaptation of this approach to extract channel networks from DEM data is straightforward and is simple both algorithmically and computationally.     

基于独立分量分析的极化SAR图像的相干斑抑制

研究基于独立分量分析( ICA)的极化合成孔径雷达（SAR）图像相干斑抑制方法。该方法将极化SAR图像斑点噪声的乘积模型,变换为应用ICA的信号独立加噪模型。并且将HV/VV的比值图像,也作为ICA的输入数据。利用ICA 的分离性,得到了分别对应于HH、HV和VV极化的三幅降噪图像。经本文方法处理后的图像,其相干斑噪声得到了有效的抑制,具有较高的等效视数,明显地改善了图像的质量。     

一种基于混沌和独立成分分析的自适应多重数字水印算法

采用混沌和独立成分分析技术,提出了一种在小波域自适应嵌入多重数字水印的算法。首先根据一个混沌标签序列从原图像中抽取一子图像,并对其进行小波分解,然后将多个互相独立的混沌水印信号依次嵌入到子图像的中频小波系数中,并自适应确定嵌入强度。根据Newman-Pearson准则检测水印是否存在,若存在,采用独立成分分析的方法提取出水印并进一步判断是否为原始水印。水印的检测与提取过程都不需要原始图像。仿真结果表明该算法对各种常见的攻击有着良好的鲁棒性,有效地解决了数字产品的多著作权问题。     

Least-squares independent component analysis

Accurately evaluating statistical independence among random variables is a key element of independent component analysis (ICA). In this letter, we employ a squared-loss variant of mutual information as an independence measure and give its estimation method. Our basic idea is to estimate the ratio of probability densities directly without going through density estimation, thereby avoiding the difficult task of density estimation. In this density ratio approach, a natural cross-validation procedure is available for hyperparameter selection. Thus, all tuning parameters such as the kernel width or the regularization parameter can be objectively optimized. This is an advantage over recently developed kernel-based independence measures and is a highly useful property in unsupervised learning problems such as ICA. Based on this novel independence measure, we develop an ICA algorithm, named least-squares independent component analysis.     

Principal independent component analysis

Conventional blind signal separation algorithms do not adopt any asymmetric information of the input sources, thus the convergence point of a single output is always unpredictable. However, in most of the applications, we are usually interested in only one or two of the source signals and prior information is almost always available. In this paper, a principal independent component analysis (PICA) concept is proposed. We try to extract the objective independent component directly without separating all the signals. A cumulant-based globally convergent algorithm is presented and simulation results are given to show the hopeful applicability of the PICA ideas.     

Contextual void patching for digital elevation models

Digital terrain models can be created by gathering a set of measurements from geometric objects. For various reasons, these models may be incomplete and thus fail to meet the requirements defined by their potential applications. In this work, we develop a novel multiresolution approach to repair the voids commonly found in digital elevation models (DEM). We use the overall shape and structure of the surrounding terrain to build a smooth patch for the void. Then, using a multiresolution approach obtained from reverse Chaikin subdivision, we extract the low-scale characteristics from the surrounding terrain and apply them to the smooth patch. The results demonstrate that our approach is effective in synthesizing models with realistic characteristics.     

Stream detection for LiDAR digital elevation models from a forested area

Airborne Laser Swath Mapping instrument technology and subsequent algorithm advances have made it possible over the last few years to map the Earth's surface and land cover at unprecedented resolution. The ability of Airborne Laser Swath Mapping technology to densely sample ground elevations beneath forest canopies is particularly important because forested watersheds have traditionally been difficult to study with remote sensing techniques. The extraction of stream networks from digital elevation models (DEMs) plays a fundamental role in modelling local and spatially distributed hydrological processes. Our approach, based on an encoding of mathematical morphological operators, is shown to systematically and accurately extract stream channel locations, forms and incipient incisions in a forested watershed. The accuracy of the method is verified using a set of error measures over simulated terrain and also over real terrain where the site was manually surveyed.     

Morphological convexity measures for terrestrial basins derived from digital elevation models

Geophysical basins of terrestrial surfaces have been quantitatively characterized through a host of indices such as topological quantities (e.g. channel bifurcation and length ratios), allometric scaling exponents (e.g. fractal dimensions), and other geomorphometric parameters (channel density, Hack's and Hurst exponents). Channel density, estimated by taking the ratio between the length of channel network (L) and the area of basin (A) in planar form, provides a quantitative index that has hitherto been related to various geomorphologically significant processes. This index, computed by taking the planar forms of channel network and its corresponding basin, is a kind of convexity measure in the two-dimensional case. Such a measure - estimated in general as a function of basin area and channel network length, where the important elevation values of the topological region within a basin and channel network are ignored - fails to capture the spatial variability between homotopic basins possessing different altitude-ranges. Two types of convexity measures that have potential to capture the terrain elevation variability are defined as the ratio of (i) length of channel network function and area of basin function and (ii) areas of basin and its convex hull functions. These two convexity measures are estimated in three data sets that include (a) synthetic basin functions, (b) fractal basin functions, and (c) realistic digital elevation models (DEMs) of two regions of peninsular Malaysia. It is proven that the proposed convexity measures are altitude-dependent and that they could capture the spatial variability across the homotopic basins of different altitudes. It is also demonstrated on terrestrial DEMs that these convexity measures possess relationships with other quantitative indexes such as fractal dimensions and complexity measures (roughness indexes).     

数字高程模型数据的无损数字水印

针对数字高程模型数据（简称DEM数据）提出广义直方图的概念,以DEM数据作为信息承载对象,采用修改广义直方图的方法嵌入水印信息,提取水印时不需要原始数据,水印提取后可完全恢复DEM数据,具有较高的经济价值。此方法计算复杂度低,适宜处理浮点精度的DEM数据,在保证较高峰值性噪比（平均42 dB）的同时嵌入信息量大（≥48.8 KB）,带水印的DEM数据能很好地保持地形形状和起伏特征。     

Topographic independent component analysis.

In ordinary independent component analysis, the components are assumed to be completely independent, and they do not necessarily have any meaningful order relationships. In practice, however, the estimated "independent" components are often not at all independent. We propose that this residual dependence structure could be used to define a topographic order for the components. In particular, a distance between two components could be defined using their higher-order correlations, and this distance could be used to create a topographic representation. Thus, we obtain a linear decomposition into approximately independent components, where the dependence of two components is approximated by the proximity of the components in the topographic representation.     

A source adaptive independent component analysis algorithm through solving the estimating equation

In this paper, a source adaptive algorithm for linear instantaneous independent component analysis is proposed. This new algorithm is based on solving the estimating equation through Newton’s method where no learning rate is needed which makes the proposed algorithm very easy to use. To achieve the source adaptivity, the density functions as well as their first and second derivatives are estimated by modified kernel density method. Empirical comparisons with several popular ICA algorithms confirm the efficiency and accuracy of the proposed algorithm.     

A common neural-network model for unsupervised exploratory data analysis and independent component analysis.

This paper presents the derivation of an unsupervised learning algorithm, which enables the identification and visualization of latent structure within ensembles of high-dimensional data. This provides a linear projection of the data onto a lower dimensional subspace to identify the characteristic structure of the observations independent latent causes. The algorithm is shown to be a very promising tool for unsupervised exploratory data analysis and data visualization. Experimental results confirm the attractiveness of this technique for exploratory data analysis and an empirical comparison is made with the recently proposed generative topographic mapping (GTM) and standard principal component analysis (PCA). Based on standard probability density models a generic nonlinearity is developed which allows both (1) identification and visualization of dichotomised clusters inherent in the observed data and (2) separation of sources with arbitrary distributions from mixtures, whose dimensionality may be greater than that of number of sources. The resulting algorithm is therefore also a generalized neural approach to independent component analysis (ICA) and it is considered to be a promising method for analysis of real-world data that will consist of sub- and super-Gaussian components such as biomedical signals.     

Intersection between spacecraft viewing vectors and digital elevation models

A method is presented for determining the latitude, longitude, and altitude of each pixel in a remotely sensed image of a planet for the case where there is significant surface topography. The method works by sequentially stepping along the line-of-sight of each pixel in an image until an intersection with the planet surface is detected. The position and altitude of each pixel on the planet can then be used for further analysis and allows comparison with other data sets. For pixels where no intersection occurs the altitude and location of the tangent point are determined. These pixels are important as they provide views of the planet limb and are useful for studying the vertical structure of the atmosphere. Provision is made for reference to an oblate spheroid when calculating the tangent point, which is required for atmospheric applications on oblate planets such as the Earth. The algorithm requires a digital elevation model, along with the viewing geometry and position of the instrument. Illustrative examples are given using the Martian MOLA topography data set for oblique and limb viewing cases.     

Coupling SAR data with forest growth models

Recent studies have shown the potential of remote sensing data at optical wavelengths to provide spatially referenced input data for process-based forest growth models. In comparison, radar remote sensing remains underexploited, despite having numerous advantages. This letter assesses the potential of radar remote sensing data to estimate tree biomass and various structural features. It is concluded that there exists a wide range of strategies for coupling radar remote sensing with forest growth models.