面向对象分类提取高分辨率多光谱影像建筑物

初步测试利用基于知识规则的面向对象分类方法从高分辨率Ikonos卫星影像上提取建筑物,包括:融合1 m全色和4 m多光谱波段影像,生成1 m分辨率的多光谱融合影像;分割融合影像;利用影像对象的光谱和空间特征执行基于对象的分类。面向对象分类提取结果与传统的基于像元最大似然分类结果进行对比,表明面向对象分类方法更适用于提取高分辨率遥感影像中的建筑物。     

Integration of high-resolution imagery and LiDAR data for object-based classification of urban area

This paper investigates the synergistic use of high-resolution multispectral imagery and Light Detection and Ranging (LiDAR) data for object-based classification of urban area. The main contribution of this paper is the development of a semi-automated object-based and rule-based classification method. In the implemented approach, the diverse knowledge about land use/land cover classes are transformed into a set of specialized rules. Further, this paper explores supervised Gaussian Mixture Models for classification, which have been primarily used for unsupervised classification. The work is carried out on test data from two different sites. Contribution of the LiDAR data resulted in a significant improvement of overall Kappa. Accuracy assessment carried out for aforementioned classification methods shows higher overall kappa for both the study sites.     

PTrees: A point-based approach to forest tree extraction from lidar data

This paper introduces PTrees, a multi-scale dynamic point cloud segmentation dedicated to forest tree extraction from lidar point clouds. The method process the point data using the raw elevation values (Z) and compute height (H = Z − ground elevation) during post-processing using an innovative procedure allowing to preserve the geometry of crown points. Multiple segmentations are done at different scales. Segmentation criteria are then applied to dynamically select the best set of apices from the tree segments extracted at the various scales. The selected set of apices is then used to generate a final segmentation. PTrees has been tested in 3 different forest types, allowing to detect 82% of the trees with under 10% of false detection rate. Future development will integrate crown profile estimation during the segmentation process in order to both maximize the detection of suppressed trees and minimize false detections.     

Evaluation of LiDAR and image segmentation based classification techniques for automatic building footprint extraction for a segment of Atlantic County,New Jersey

Extracting high-quality building footprints is a basic requirement in multiple sectors of town planning, disaster management, 3D visualization, etc. In the current study, we compare three different techniques for acquiring building footprints using (i) LiDAR, (ii) object-oriented classification (OOC) applied on high-resolution aerial photographs and (iii) digital surface models generated from interpolated LiDAR point cloud data. The three outputs were compared with a digitized sample of building polygons quantitatively by computing the errors of commission and omission, and qualitatively using statistical operations. These findings showed that building footprints derived from OOC gave highest regression and correlation values with least commission error. The R² and R values (0.86 and 0.92, respectively) imply that the footprint areas derived by OOC matched more closely with the actual area of buildings, while a low commission error of 24.7% represented a higher number of footprints as correctly classified.     

Development of a modified neural network-based land cover classification system using automated data selector and multiresolution remotely sensed data

Integrating multiple images with artificial neural networks (ANN) improves classification accuracy. ANN performance is sensitive to training datasets. Complexity and errors compound when merging multiple data, pointing to needs for new techniques. Kohonen's self-organizing mapping (KSOM) neural network was adapted as an automated data selector (ADS) to replace manual training data processes. The multilayer perceptron (MLP) network was then trained using automatically extracted datasets and used for classification. Two hypotheses were tested: ADS adapted from the KSOM network provides adequate and reliable training datasets, improving MLP classification performance; and fusion of Landsat thematic mapper (TM) and SPOT images using the modified ANN approach increases accuracy. ADS adapted from the KSOM network improved training data quality and increased classification accuracy and efficiency. Fusion of compatible multiple data can improve performance if appropriate training datasets are collected. This proved to be a viable classification scheme particularly where acquiring sufficient and reliable training datasets is difficult.     

Generalization of DEM for terrain analysis using a compound method

This paper reports an investigation into the generalization of a grid-based digital elevation model (DEM) for the purpose of terrain analysis. The focus is on the method of restructuring the grid-based surface elevation data to form a triangulated irregular network (TIN) that is optimized to keep the important terrain features and slope morphology with the minimum number of sample points. The critical points of the terrain surface are extracted from the DEM based on their significance, measured not only by their local relief, but also by their importance in identifying inherent geomorphological and drainage features in the DEM. A compound method is proposed by integrating the traditional point-additive and feature-point methods to construct a drainage-constrained TIN. The outcome is then compared with those derived from other selected methods including filtering, point-additive or feature-point algorithms. The results show that the compound approach is capable of taking advantage of both point-additive and feature-point algorithms to maximally keep the terrain features and to maintain RMSE at an acceptable level, while reducing the elevation data points by over 99%. The analytical result also shows that the proposed method outperforms the compared methods with better control in retaining drainage features at the same level of RMSE.     

A heuristic multi-criteria classification approach incorporating data quality information for choropleth mapping

ABSTRACT

Despite conceptual and technology advancements in cartography over the decades, choropleth map design and classification fail to address a fundamental issue: estimates that are statistically indifferent may be assigned to different classes on maps or vice versa. Recently, the class separability concept was introduced as a map classification criterion to evaluate the likelihood that estimates in two classes are statistical different. Unfortunately, choropleth maps created according to the separability criterion usually have highly unbalanced classes. To produce reasonably separable but more balanced classes, we propose a heuristic classification approach to consider not just the class separability criterion but also other classification criteria such as evenness and intra-class variability. A geovisual-analytic package was developed to support the heuristic mapping process to evaluate the trade-off between relevant criteria and to select the most preferable classification. Class break values can be adjusted to improve the performance of a classification.     

Object-based classification of hyperspectral data using Random Forest algorithm

This paper presents a new framework for object-based classification of high-resolution hyperspectral data. This multi-step framework is based on multi-resolution segmentation (MRS) and Random Forest classifier (RFC) algorithms. The first step is to determine of weights of the input features while using the object-based approach with MRS to processing such images. Given the high number of input features, an automatic method is needed for estimation of this parameter. Moreover, we used the Variable Importance (VI), one of the outputs of the RFC, to determine the importance of each image band. Then, based on this parameter and other required parameters, the image is segmented into some homogenous regions. Finally, the RFC is carried out based on the characteristics of segments for converting them into meaningful objects. The proposed method, as well as, the conventional pixel-based RFC and Support Vector Machine (SVM) method was applied to three different hyperspectral data-sets with various spectral and spatial characteristics. These data were acquired by the HyMap, the Airborne Prism Experiment (APEX), and the Compact Airborne Spectrographic Imager (CASI) hyperspectral sensors. The experimental results show that the proposed method is more consistent for land cover mapping in various areas. The overall classification accuracy (OA), obtained by the proposed method was 95.48, 86.57, and 84.29% for the HyMap, the APEX, and the CASI data-sets, respectively. Moreover, this method showed better efficiency in comparison to the spectral-based classifications because the OAs of the proposed method was 5.67 and 3.75% higher than the conventional RFC and SVM classifiers, respectively.     

一种计算机辅助设计数据与GIS数据转换方法

针对GIS数据与计算机辅助设计(CAD)数据之间的差异性,以及现有转换方法存在的问题,该文提出了一种计算机辅助设计数据处理方法——采用空间数据操作引擎(FME)和Arc Map联合使用的方法。该方法能够较好地解决计算机辅助设计数据向GIS数据转换的难题,实现了对计算机辅助设计数据的批量转换和处理。以某市的地形图为数据源进行实例论证,实验结果显示运用该文所提方法能够更好地处理计算机辅助设计数据,为计算机辅助设计地形图的转换提供了可行的方案;特别是对转换出来的数据进行了拓扑重建、附属性值以及多边形数据分类处理,能够更好地保持数据的精度和完整性,更有利于数据的入库。     

极化SAR参数优化与光学波谱相结合的面向对象土地覆盖分类

提出一种基于极化参数优化的面向对象分类方法。该方法结合光学和SAR数据,有效提高了对地物的识别能力。本文方法的关键在于:在■分解中,使用光学影像指导SAR影像选择同质点,使其更精确地估计极化参数并结合光学波谱信息作为输入特征;使用面向对象的分类方法,仅将光学影像作为分割输入,避免SAR噪声引起的分割错误。以美国Bakersfield地区的Sentinel-1/2数据为例,确定7种地物类型,对比分析不同输入与不同分类器对分类结果的影响。研究表明,优化输入参数在纹理丰富区域能够有效提高分类精度;面向对象的分类结果更加稳定并较好地维持地表几何特征;改进分类方法较传统分类方法总体精度提高了近10%,达到92.6%。     

Integration of WorldView-2 and airborne LiDAR data for tree species level carbon stock mapping in Kayar Khola watershed,Nepal

Integration of WorldView-2 satellite image with small footprint airborne LiDAR data for estimation of tree carbon at species level has been investigated in tropical forests of Nepal. This research aims to quantify and map carbon stock for dominant tree species in Chitwan district of central Nepal. Object based image analysis and supervised nearest neighbor classification methods were deployed for tree canopy retrieval and species level classification respectively. Initially, six dominant tree species (Shorea robusta, Schima wallichii, Lagerstroemia parviflora, Terminalia tomentosa, Mallotus philippinensis and Semecarpus anacardium) were able to be identified and mapped through image classification. The result showed a 76% accuracy of segmentation and 1970.99 as best average separability. Tree canopy height model (CHM) was extracted based on LiDAR’s first and last return from an entire study area. On average, a significant correlation coefficient (r) between canopy projection area (CPA) and carbon; height and carbon; and CPA and height were obtained as 0.73, 0.76 and 0.63, respectively for correctly detected trees. Carbon stock model validation results showed regression models being able to explain up to 94%, 78%, 76%, 84% and 78% of variations in carbon estimation for the following tree species: S. robusta, L. parviflora, T. tomentosa, S. wallichii and others (combination of rest tree species).     

Improvement of MEMS-IMU/GPS performance using fuzzy modeling

Performance improvement of integrated Inertial Measurement Units (IMU) utilizing micro-electro-mechanical-sensors (MEMS) and GPS is described in this paper. An offline pre-defined Fuzzy model is employed to improve the system performance. The Fuzzy model is used to predict the position and velocity errors, which are the inputs to a Kalman Filter (KF) during GPS signal outages. The proposed model has been verified on real MEMS inertial data collected in a land vehicle test. A number of 30-s GPS outages were simulated during the data processing at different times and under different vehicle dynamics. Performance of the suggested Fuzzy model was compared to that of the traditional KF particularly during the simulated GPS outages. The test results indicate that the proposed Fuzzy model can efficiently compensate for GPS updates during short outages.     

Quantifying shoreline change in Funafuti Atoll,Tuvalu using a time series of Quickbird,Worldview and Landsat data

Funafuti Atoll, Tuvalu is located in the southwestern Pacific Ocean, which has experienced some of the highest rates of global sea-level rise over the past 60 years. Atoll islands are low-lying accumulations of reef-derived sediment that provide the only habitable land in Tuvalu, and are considered vulnerable to the myriad possible impacts of climate change, especially sea-level rise. This study examines the shoreline change of twenty-eight islands in Funafuti Atoll between 2005 and 2015 using 0.65 m QuickBird, 0.46 m WorldView-2, and 0.31 m WorldView-3 imagery using an image segmentation and decision tree classification. Shoreline change estimates are compared to previous study that used a visual interpretation approach. The feasibility of estimating island area with Landsat-8 Operational Land Imager (OLI) data is explored using CLASlite software. Results indicate a 0.13% (0.35 ha) decrease in net island area over the study time period, with 13 islands decreasing in area and 15 islands increasing in area. Substantial decreases in island area occurred on the islands of Fuagea, Tefala and Vasafua, which coincides with the timing of Cyclone Pam in March, 2015. Comparison between the WorldView-2 shoreline maps and those created from Landstat-8 indicate that the estimates tend to be in higher agreement for islands that have an area > 0.5 ha, a compact shape, and no built structures. Ten islands had > 90% agreement, with percent disagreements ranging from 2.78 to 100%. The methods and results of this study speak to the potential of automated EoV shoreline monitoring through segmentation and classification tree approach, which would reduce down data processing and analysis time. With the growing constellation of high and medium spatial resolution satellite-based sensors and the development of semi or fully automated image processing technology, it is now possible to remotely assess the short and medium-term shoreline dynamics on dynamic atolls. Landsat estimates were reasonably matched to those derived from fine resolution imagery, with some caveats about island size and shape.     

遥感数据模糊不确定性来源及其处理方法的探讨

近年来遥感信息的应用越加广泛,数据不确定性表现也较明显。本文主要是对遥感数据的质量也就要作进一步的分析:遥感信息本身固有的不确定性,遥感数据的分析与处理过程中所引入的不确定性。在总结不确定性存在原因的基础之上,对不确定性的处理方法作进一步的归纳。     

遥感图像土地覆盖分类中多源特征数据选择研究

多源特征数据可以提高遥感图像的分类精度,选择合适的特征数据十分重要。利用基尼指数对多尺度纹理信息、主成分变换前三分量、地形数据等特征进行选择,选出最佳特征子集。利用支持向量机、神经网络分类法、最大似然法分别对全部特征数据和最佳特征子集结合多光谱数据进行分类。实验结果表明:基尼指数可以有效地对多源特征数据进行选择,特征选择可以提高分类器效率,提高分类精度。     

改进的P-SVM支持向量机与遥感数据分类

本文介绍了将P-SVM算法引入多光谱/高分辨率遥感数据的分类, 并且展示了卫星ASTER和航空ADS40数字影像分类的技术过程和结果验证。结果表明：P-SVM方法的分类精度不低于SVM, 并减少了时耗。     

一种面向对象的遥感影像城市绿地提取方法

针对目前面向对象方法在高分辨率遥感影像中提取绿地专题信息的特点,以及受到城市区域范围尺度限制因素的影响,本文提出一种利用高分辨率遥感影像提取城市区域范围绿地专题信息的方法,结合影像分割、影像面向对象模板统计和基于多阀值的模糊分类技术的优势,林地、密草地和疏草地专题信息被提取出来.研究中提出利用基于面向对象多尺度分割算法...     

Curve matching approaches to waveform classification: a case study using ICESat data

Light Detection and Ranging (LiDAR) waveforms are being increasingly used in many forest and urban applications, especially for ground feature classification. However, most studies relied on either discretizing waveforms to multiple returns or extracting shape metrics from waveforms. The direct use of the full waveform, which contains the most comprehensive and accurate information has been scarcely explored. We proposed to utilize the complete waveform to test its ability to differentiate between objects having distinct vertical structures using curve matching approaches. Two groups of curve matching approaches were developed by extending methods originally designed for pixel-based hyperspectral image classification and object-based high spatial image classification. The first group is based on measuring the curve similarity between an unknown waveform and a reference waveform, including curve root sum squared differential area (CRSSDA), curve angle mapper (CAM), and Kullback–Leibler (KL) divergence. The second group assesses the curve similarity between an unknown and reference cumulative distribution functions (CDFs) of their waveforms, including cumulative curve root sum squared differential area (CCRSSDA), cumulative curve angle mapper (CCAM), and Kolmogorov–Smirnov (KS) distance. When employed to classify open space, trees, and buildings using ICESat waveform data, KL provided the highest average classification accuracy (87%), closely followed by CCRSSDA and CCAM, and they all significantly outperformed KS, CRSSDA, and CAM based on 15 randomized sample sets.     

Discrimination of species composition types of a grazed pasture landscape using Sentinel-1 and Sentinel-2 data

Species composition is one of the important measurable indices of alpha diversity and hence aligns with the measurable Essential Biodiversity Variables meant to fulfil the Aichi Biodiversity Targets by 2020. Graziers also seek for pasture fields with varied species composition for their livestock, but visual determination of the species composition is not practicable for graziers with large fields. Consequently, this study demonstrated the capability of Sentinel-1 Synthetic Aperture Radar (S1) and Sentinel-2 Multispectral Instrument (S2) to discriminate pasture fields with single-species composition, two-species composition and multi-species composition for a pastoral landscape in Australia. The study used K-Nearest Neighbours (KNN), Random Forest (RF) and Support Vector Machine (SVM) classifiers to evaluate the strengths of S1-alone and S2-alone features and the combination of these S1 and S2 features to discriminate the composition types. For the S1 experiment, KNN which was the reference classifier achieved an overall accuracy of 0.85 while RF and SVM produced 0.74 and 0.89, respectively. The S2 experiment produced accuracies higher than the S1 in that the overall performance of the KNN classifier was 0.87 while RF and SVM were 0.93 and 0.89, respectively. The combination of the S1 and S2 features elicited the highest accuracy estimates of the classifiers in that the KNN classifier recorded 0.89 while RF and SVM produced 0.96 and 0.93, respectively. In conclusion, the inclusion of S1 features improve the classifiers created with S2 features only.