城区航空影像中铅垂线的自动提取

为了将铅垂线应用于城区空中三角测量,通过对大比例尺航空影像中铅垂线特点的分析,提出了一种有效的铅垂线提取算法,即首先对航空影像进行必要的预处理并自动提取直线,然后使用累积法从提取直线中分离出铅垂线并确定像底点坐标,最后应用共点约束的最小二乘直线模板匹配算法进一步提高铅垂线的提取精度及像底点的交会精度。     

一种有效的航拍图像中直线提取算法

首先对航拍图像进行小尺度高斯滤波;然后进行边缘检测,得到边缘检测幅度图像;再在边缘检测幅度图像中进行启发式搜索连接,提取出符合直线模型的直线搜索轨迹;最后根据直线搜索轨迹的判别获取直线。实验结果表明,该算法可以在航拍图像中提取出真实的直线,是一种有效的直线提取算法。     

一种新的边缘直线拟合方法

针对直线提取过程中的断裂问题,该文提出一种新的边缘拟合直线方法,首先采用Canny算法提取边缘图;接着利用改进ZS算法对边缘图进行细化处理;然后通过端点检测、交叉点跟踪、闭合边缘跟踪等方法,实现边缘链码的完整性提取;最后利用边缘链码跟踪顺序进行直线提取。通过多幅影像的实验结果表明,该文所提出的算法能够精确实现边缘直线信息的拟合,便于对影像结构信息的清晰理解。     

基于自适应中值滤波和改进Hough变换的直线提取

直线是遥感图像的重要纹理特征,直线提取是直线匹配、遥感影像数据的三维重建、图像配准等的前提基础。本文针对传统Hough变换存在的问题以及MatLab软件工具箱中现有滤波算法的不足,提出对Hough变换进行改进,设计自适应中值滤波,并结合Canny边缘检测算子实现了直线的准确提取,使得直线提取算法具有较强的抗噪性能,不误检直线。     

基于EDLines的遥感影像直线快速提取方法研究

随着现代科学技术的发展,大数据已经成为这个时代的标志。因此快速、准确地采集图像数据变得十分重要。本文研究的是实时线段检测算法EDLines,首先通过实验对比分析及所给公式,可以获得梯度幅度和方向、线验证参数、梯度阈值、锚点阈值和扫描间隔参数值,基于所获得的参数值,在处理不同类型图像时不需要改变其参数就能获取准确的结果,然后对所研究的算法EDLines与现今已知运行较快的直线提取算法进行对比分析,得到本文所研究的EDLines算法相比较其他直线提取算法都具有最优性。最后,根据本文研究的EDLines算法对遥感影像进行直线提取,获得清晰且假阳性较少的直线提取图像。超高的精度和计算速度使实时计算机视觉和图像处理的应用迈出了崭新的一步。     

Freeman链码优先级直线提取算法研究

针对目前直线提取算法的局限性,提出了一种简单高效的提取图像中目标物体直线边界的算法。本算法基于链码思想和数字直线特征,首先对图像进行边缘检测,对边缘检测后的图像按照优先级进行链码跟踪,获取初始链码;然后通过距离约束提高链码的直线性,剔除噪声和细节;最后进行直线合并,完成直线提取。实验表明,本文提出的算法能高效、准确地检测出图像中物体边界的直线,特别对较大的影像计算量小,抗噪能力强,适用于实时处理。     

多源高分辨率遥感影像自动匹配算法

提出一种基于同名直线约束的多源高分辨率遥感影像自动匹配方法。对影像建立高斯滤波模型去噪,然后进行边缘检测,在边缘影像上进行Hough变换,通过端点检测的方法来获取直线;通过粗匹配参数进行同名直线粗匹配,应用梯度模型对同名直线进行精匹配;对影像提取特征点,特征点经过同名直线约束后,在对应影像上利用同名直线约束建立匹配范围。采用最小欧式距离准则在给定范围内提取初始同名点;利用RANSAC算法剔除错误同名点对,以获取最终的匹配结果。实验结果表明,与传统的SIFT匹配算法比较,方法具有可靠性好、提取同名点数量多的优点。     

一种高分辨率光学卫星遥感影像直线匹配算法

高分辨率光学卫星遥感影像直线自动匹配算法的具体流程包括:进行影像尺度调整,对影像进行边缘提取,采用Hough变换提取直线;同时基于直线几何特征和灰度特征提取同名直线,利用Hough变换模型,将线转换为参数空间中的点,并建立转换模型;最后采用RANSAC算法剔除错误同名直线,以获取最终的匹配结果。实验结果表明,本文提出的算法能显著提高影像同名直线提取的数量和精度,特别是对异源影像具有抗噪能力强、鲁棒性好等特点。     

基于改进RANSAC算法的道路直线提取方法

直线检测是计算机视觉领域的一项重要内容,也是遥感影像信息提取的基本过程。随机抽样一致性算法(RANSAC)常用来进行包括直线在内的目标提取,是在计算机视觉领域应用较广泛的估计算法之一,但其计算效率较低。因此提出了一种基于序贯概率的改进型多目标RANSAC算法,在利用边缘检测限定随机抽样原始样本集的基础上,运用该算法提取了高分辨率航空影像数据中的道路边界线,大大提升了计算效率。     

基于多通道梯度特征的彩色图像直线提取方法

经典LSD直线提取算法仅采用图像的灰度信息,未充分考虑R,G,B彩色通道中的梯度差异,导致部分重要直线检测率低且连续性较差的问题,文中提出基于多通道梯度特征的彩色图像直线提取方法。首先计算彩色图像的R,G,B通道的梯度值及其对应的灰度图像的梯度值,然后取梯度的最大值及其对应的梯度方向作为初始梯度值,最后通过区域生长和直线拟合完成直线检测。采用多幅York城市图像数据库中的图像和真实影像进行实验,结果表明文中方法在彩色图像中的直线提取效果优于经典LSD直线提取算法,提取结果中直线信息更丰富,连续性更好。     

利用线阵卫星立体像对的平行特性生成核线影像

针对线阵推扫卫星影像的核线特点,研究了基于平行特征的核线影像生成方法。首先将影像分块并利用有理函数模型和投影轨迹法计算核线对,在分析核线斜率及直线拟合误差的基础上研究核线的平行特征,平行性好时旋转核线倾斜角相同的角度生成核线影像,最后以同名点上下视差为标准评价方法的可行性。试验结果表明,基于平行特性直接旋转一定角度的核线影像生成方法虽不具备普遍适用性,但对WorldView-2卫星4波段多光谱和全色等特定影像是可行的。     

Mapping trees in high resolution imagery across large areas using locally variable thresholds guided by medium resolution tree maps

Large area tree maps, important for environmental monitoring and natural resource management, are often based on medium resolution satellite imagery. These data have difficulty in detecting trees in fragmented woodlands, and have significant omission errors in modified agricultural areas. High resolution imagery can better detect these trees, however, as most high resolution imagery is not normalised it is difficult to automate a tree classification method over large areas. The method developed here used an existing medium resolution map derived from either Landsat or SPOT5 satellite imagery to guide the classification of the high resolution imagery. It selected a spatially-variable threshold on the green band, calculated based on the spatially-variable percentage of trees in the existing map of tree cover. The green band proved more consistent at classifying trees across different images than several common band combinations. The method was tested on 0.5 m resolution imagery from airborne digital sensor (ADS) imagery across New South Wales (NSW), Australia using both Landsat and SPOT5 derived tree maps to guide the threshold selection. Accuracy was assessed across 6 large image mosaics revealing a more accurate result when the more accurate tree map from SPOT5 imagery was used. The resulting maps achieved an overall accuracy with 95% confidence intervals of 93% (90–95%), while the overall accuracy of the previous SPOT5 tree map was 87% (86–89%). The method reduced omission errors by mapping more scattered trees, although it did increase commission errors caused by dark pixels from water, building shadows, topographic shadows, and some soils and crops. The method allows trees to be automatically mapped at 5 m resolution from high resolution imagery, provided a medium resolution tree map already exists.     

Use of Kendall's coefficient of concordance to assess agreement among observers of very high resolution imagery

Ground-based vegetation monitoring methods are expensive, time-consuming and limited in sample size. Aerial imagery is appealing to managers because of the reduced time and expense and the increase in sample size. One challenge of aerial imagery is detecting differences among observers of the same imagery. Six observers analysed a set of 1-mm ground sample distance aerial imagery for graminoid species composition and important ground-cover characteristics. Kendall's coefficient of concordance (W) was used to measure agreement among observers. The group of six observers was concordant when assessed as a group. When each of the observers was assessed independently against the other five, lack of agreement was found for those graminoid species that were difficult to identify in the aerial images.     

基于整体变分模型的影像阴影检测算法研究

将影像上普遍存在的阴影视为图像退化的一种特殊形式，以整体变分模型为基础，以影像上阴影区域亮度普遍较暗且较均匀、阴影区域和非阴影区域之间的反差普遍较大的特点为约束，导出了整体变分模型用于影像上阴影检测的基本算法。通过对若干幅实际影像的阴影检测实验表明，本文算法对灰度影像和彩色影像上阴影区域的检测是有效的。     

Interpretation of Synthetic Aperture Radar (SAR) imagery for geological appraisal: A comparative study in Anantapur district of Andhra Pradesh

The paper describes the details of a comparative study of geological interpretations carried out from Synthetic Aperture Radar (SAR) imagery, Landsat MSS (B & W) imagery and Aerial Photographs, covering 2100 sq km of area in Anantapur district of Andhra Pradesh. The area comprises Peninsular—Gneissic Complex and rocks of Dharwar and Cuddapah Super Groups beside the Quaternary alluvial deposits along the Penneru river and its tributaries. Geomorphologically the areas is represented by denudational, fluvial and structural landforms. The study indicates that the details of the geological and geomorphological maps prepared from SAR imagery and aerial photographs are comparable despite the smaller scale of SAR imagery while the same are not exhibited in Landsat imagery mainly due to its low resolution. Although broad lithological units are possible to be discriminated on SAR as well as aerial photographs, some of the finer rock types viz. gabbroic dykes could be discriminated from the delerite dykes in the SAR imagery due to their different surface roughness. Stereoscopic coverage and enhanced micro-relief of SAR imagery gives better geomorphological details in comparison to aerial photographs. A detailed study of lineaments has also been carried out which shows that in SAR imagery there is over-representation of short lineaments due to enhanced micro-relief and radarshadow effects across the look direction and under-representation of lineaments along the look direction. Landsat imagery is perhaps the best for demarcating lineaments of regional magnitude while aerial photographs are good for depicting shorter lineaments. However, certain lineaments seen in SAR imagery are often not continuously seen on aerial photographs.     

Parallelizing maximum likelihood classification on computer cluster and graphics processing unit for supervised image classification

ABSTRACT

Supervised image classification has been widely utilized in a variety of remote sensing applications. When large volume of satellite imagery data and aerial photos are increasingly available, high-performance image processing solutions are required to handle large scale of data. This paper introduces how maximum likelihood classification approach is parallelized for implementation on a computer cluster and a graphics processing unit to achieve high performance when processing big imagery data. The solution is scalable and satisfies the need of change detection, object identification, and exploratory analysis on large-scale high-resolution imagery data in remote sensing applications.     

THE CONCEPT OF AN INTEGRATED REPOSITORY OF SPACE IMAGERY AND NEW CRITERIA FOR ITS GEOGRAPHIC AND CARTOGRAPHIC ANALYSIS

This paper outlines some principles believed necessary for the establishment of integrated collections of remote sensing imagery, including, at the national level, a unified state image repository for the USSR. A multi-criterion classification of imagery is introduced, which provides a framework for structuring such a repository, and a number of measures for evaluating the utility of imagery within it are described. One of the latter is “geographic” resolution (levels of image detail), which provides a relatively straightforward indication of the type of geographic information embedded within particular types of remote sensing imagery. Translated by Jay K. Mitchell, PlanEcon, Inc., Washington, DC 20005 from: Vestnik Moskovskogo Universiteta, geografiya, 1988, No. 6, pp. 53-62.     

Comparative study on projected clustering methods for hyperspectral imagery classification

In this study, projected clustering is introduced to hyperspectral imagery for unsupervised classification. The main advantage of projected clustering lies in its ability to simultaneously perform feature selection and clustering. This framework also allows selection of different sets of dimensions (features/bands) for different clusters. This framework provides an effective way to address the issues associated with the high dimensionality of the data. Experiments are conducted on both synthetic and real hyperspectral imagery. For this purpose, projected clustering algorithms are implemented and compared with k-means and k-means preceded by principal component analysis. Preliminary analyses of studied algorithms on synthetic hyperspectral imagery demonstrate good results. For real hyperspectral imagery, only ORCLUS is able to produce acceptable results as compared to other unsupervised methods. The main concern lies with identification of right parameter settings. More experiments are required in this direction.     

Progress in the Development of a High Performance Airborne Digital Sensor

Joint development work by LH Systems and Deutsches Zentrum für Luft- und Raumfahrt (German Aerospace Center) has produced encouraging results using forward, nadir and backward looking linear arrays on the focal plane to provide panchromatic imagery and geometric information, supplemented by further arrays to acquire multispectral imagery suitable for both high precision photogrammetric mapping and image processing for interpretative purposes. The geometric characteristics of line scanner imagery necessitate line-by-line rectification for aircraft tilts and shifts. Satisfactory execution of this process is enhanced by using supplementary data from high performance, on-board GPS and inertial measurement systems. Similarly, high demands are placed on other sub-systems, such as the camera mount, lens, electronics and storage technology. In addition to rectification for aircraft tilts and shifts, rectification for terrain characteristics is also required in order to generate colour and false colour composite images, since the various multispectral arrays are in different places on the focal plane. The special geometry affects triangulation. Thereafter, the imagery can be processed using existing software packages from both photogrammetry and remote sensing.
The concept has been demonstrated in several successful test flights and the production model is scheduled for market introduction at the ISPRS Congress in July 2000. The imagery from the new sensor will fulfil many market requirements between the highest resolution film imagery (<0.1 m) and high resolution space imagery (1m to 10 m). The sensor's unique blend of multispectral information with high quality geometric information will give rise to numerous new applications.