Performance evaluation of classification trees for building detection from aerial images and LiDAR data: a comparison of classification trees models

This study assesses the performance of three classification trees (CT) models (entropy, gain ratio and gini) for building detection by the fusion of airborne laser scanner data and multispectral aerial images. Data from four study areas with different sensors and scene characteristics were used to assess the performance of the models. The process of performance evaluation is based on four criteria: model validation and testing, classification accuracies, relative importance of input variables, as well as transferability of CT derived from one data set to another. The LiDAR point clouds were filtered to generate a digital terrain model (DTM) based on the orthogonal polynomials, and then a digital surface model (DSM) and the normalized digital surface model (nDSM) were generated. A set of 26 uncorrelated feature attributes were derived from the original aerial images, LiDAR intensity image, DSM and nDSM. Finally, the three CT models were used to classify buildings, trees, roads and ground from aerial images, LiDAR data and the generated attributes, with the most accurate average classifications of 95% being achieved. The entropy splitting algorithm proved to be a preferable algorithm for building detection from aerial images and LiDAR data.     

A model-based approach for automatic building database updating from high-resolution space imagery

This article presents an approach for automatic building database updating from high-resolution space imagery based on the support vector machine (SVM) classification and building models. The developed approach relies on an idea that the buildings are similar in shape within an urban block or a neighbourhood unit. First, the building patches are detected through classification of the pan-sharpened high-resolution space imagery along with the normalized digital surface model (nDSM) and the normalized difference vegetation index (NDVI) using the binary SVM classifier. Then, the buildings that exist in the vector database but not seen in the image are detected through the analyses of the detected building patches. The buildings, which were constructed after the compilation date of the existing vector database, are extracted through the proposed model-based approach that utilizes the existing building database as a building model library. The approach was implemented in selected urban areas of the Batikent district of Ankara, the capital city of Turkey, using the IKONOS images and the existing building database. The results validated the success of the developed approach, with the building extraction accuracy computed to be higher than 80%.     

Use of lidar-derived NDTI and intensity for rule-based object-oriented extraction of building footprints

Buildings play an essential role in urban intra-construction, planning, and climate. The precise knowledge of building footprints not only serves as a primary source for interpreting complex urban characteristics, but also provides regional planners with more realistic and multidimensional scenarios for urban management. The recently developed airborne light detection and ranging (lidar) technology provides a very promising alternative for building-footprint measurement. In this study, lidar intensity data, a normalized digital surface model (nDSM) of the first and last returns, and the normalized difference tree index (NDTI) derived from the two returns are used to extract building footprints using rule-based object-oriented classification. The study area is chosen in London, Ontario, based on the various types of buildings surrounded by trees. An integrated segmentation approach and a hierarchical rule-based classification strategy are proposed during the process. The results indicate that the proposed object-based classification is a very effective semi-automatic method for building-footprint extraction, with buildings and trees successfully separated. An overall accuracy of 94.0% and a commission error of 6.3% with a kappa value of 0.84 are achieved. Lidar-derived NDTI and intensity data are of great importance in object-based building extraction, and the kappa value of the proposed method is double that of the object-based method without NDTI or intensity.     

Automated building extraction from IKONOS images in suburban areas

The article addresses automatic building extraction from IKONOS images in suburban areas. In the proposed approach, we used a stereo pair of IKONOS images. Automatic photogrammetric methods of image matching were used to generate a digital surface model (DSM) and a digital elevation model. In further processing, single-image methods were used. The orthophotos of individual bands were created. The initial building mask was generated from the calculated normalized DSM (nDSM). The calculated normalized difference vegetation index and the road data extracted from the existing topographical database were used to remove vegetation and traffic surfaces. The mask was further improved with our own combination of methods based on non-linear diffusion filtering, unsupervised classification, colour segmentation and region growing. The final mask was vectorized using the Hough transform. Compared with a reference building database, 83.2% of the buildings in the test area were detected using the proposed approach with a quality percentage (how likely a building pixel produced by an automatic approach is correct) of 49.46.     

A hybrid data fusion approach to the hierarchical mapping of land-cover types in the urban core

It is difficult to map land covers in the urban core due to the close proximity of high-rise buildings. This difficulty is overcome with a proposed hybrid, the hierarchical method via fusing PAN-sharpened WorldView-2 imagery with light detection and ranging (lidar) data for central Auckland, New Zealand, in two stages. After all features were categorized into ‘ground’ and ‘above-ground’ using lidar data, ground features were classified from the satellite data using the object-oriented method. Above-ground covers were grouped into four types from lidar-derived digital surface model (nDSM) based on rules. Ground and above-ground features were classified at an accuracy of 94.1% (kappa coefficient or κ = 0.913) and 93.7% (κ = 0.873), respectively. After the two results were merged, the nine covers achieved an overall accuracy of 93.7% (κ = 0.902). This accuracy is highly comparable to those reported in the literature, but was achieved at much less computational expense and complexity owing to the hybrid workflow that optimizes the efficiency of the respective classifiers. This hybrid method of classification is robust and applicable to other scenes without modification as the required parameters are derived automatically from the data to be classified. It is also flexible in incorporating user-defined rules targeting hard-to-discriminate covers. Mapping accuracy from the fused complementary data sets was adversely affected by shadows in the satellite image and the differential acquisition time of imagery and lidar data.     

Generation of a DTM and building detection based on an MPF through integrating airborne lidar data and aerial images

This study presents an approach that uses airborne light detection and ranging (lidar) data and aerial imagery for creating a digital terrain model (DTM) and for extracting building objects. The process of creating the DTM from lidar data requires four steps in this study: pre-processing, segmentation, extraction of ground points, and refinement. In the pre-processing step, raw data are transformed to raster data. For segmentation, we propose a new mean planar filter (MPF) that uses a 3 × 3 kernel to divide lidar data into planar and nonplanar surfaces. For extraction of ground points, a new method to extract additional ground points in forest areas is used, thus improving the accuracy of the DTM. The refinement process further increases the accuracy of the DTM by repeated comparison of a temporary DTM and the digital surface model. After the DTM is generated, building objects are extracted via a proposed three-step process: detection of high objects, removal of forest areas, and removal of small areas. High objects are extracted using the height threshold from the normalized digital surface model. To remove forest areas from among the high objects, an aerial image and normalized digital surface model from the lidar data are used in a supervised classification. Finally, an area-based filter eliminates small areas, such as noise, thus extracting building objects. To evaluate the proposed method, we applied this and three other methods to five sites in different environments. The experiment showed that the proposed method leads to a notable increase in accuracy over three other methods when compared with the in situ reference data.     

Pixel- and feature-level fusion of hyperspectral and lidar data for urban land-use classification

The complexity of urban areas makes it difficult for single-source remotely sensed data to meet all urban application requirements. Airborne light detection and ranging (lidar) can provide precise horizontal and vertical point cloud data, while hyperspectral images can provide hundreds of narrow spectral bands which are sensitive to subtle differences in surface materials. The main objectives of this study are to explore: (1) the performance of fused lidar and hyperspectral data for urban land-use classification, especially the contribution of lidar intensity and height information for land-use classification in shadow areas; and (2) the efficiency of combined pixel- and object-based classifiers for urban land-use classification. Support vector machine (SVM), maximum likelihood classification (MLC), and object-based classifiers were used to classify lidar, hyperspectral data and their derived features, such as the normalized digital surface model (nDSM), normalized difference vegetation index (NDVI), and texture measures, into 15 urban land-use classes. Spatial attributes and rules were used to minimize misclassification of the objects showing similar spectral properties, and accuracy assessments were carried out for the classification results. Compared with hyperspectral data alone, hyperspectral–lidar data fusion improved overall accuracy by 6.8% (from 81.7 to 88.5%) when the SVM classifier was used. Meanwhile, compared with SVM alone, the combined SVM and object-based method improved OA by 7.1% (from 87.6 to 94.7%). The results suggest that hyperspectral–lidar data fusion is effective for urban land-use classification, and the proposed combined pixel- and object-based classifiers are very efficient and flexible for the fusion of hyperspectral and lidar data.     

Automatic detection of building roofs from point clouds produced by the dense image matching technique

Three-dimensional (3D) spatial information of object points is a vital requirement for many disciplines. Laser scanning technology and techniques based on image matching have been used extensively to produce 3D dense point clouds. These data are used frequently in various applications, such as the generation of digital surface model (DSM)/digital terrain model (DTM), extracting objects (e.g., buildings, trees, and roads), 3D modelling, and detecting changes. The aim of this study was to extract the building roof points automatically from the 3D point cloud data created via the image matching techniques with optical aerial images (with red, green, and blue band (RGB) and infrared (IR)). In the first stage of the study, as an alternative to laser scanning technology, which is more expensive than optical imaging systems, the 3D point clouds were produced by matching high-resolution images using a Semi Global Matching algorithm. The normalized difference vegetation index (NDVI) values for each point were calculated using the spectral information (RGB + IR) in the 3D point cloud data, and the points that represented the vegetation cover were determined using these values. In the second stage, existing ground and non-ground points that were free of vegetation cover were determined within the point cloud. Subsequently, only the points on the roof of the building were detected automatically using the proposed algorithm. Thus, points of the roofs of buildings located in areas with different topographic characteristics were detected automatically detected using only images. It was determined that the average values of correctness (Corr), completeness (Comp), and quality (Q) of the pixel-based accuracy analysis metrics were 95%, 98%, and 93%, respectively, in the selected test areas. According to the results of the accuracy analysis, it is clear that the proposed algorithm is very successful in automatic extraction of building roof points.     

True orthoimage generation in urban areas with very tall buildings

The orthoimage usually serves as a valuable base layer in GIS. With an increasing demand in many urban GIS applications, orthoimages in urban areas are required to represent spatial objects in their true positions. However, the traditional methods for orthoimage generation did not consider features (e.g. occlusion, shadow, etc.) of spatial objects (e.g. bridges and buildings), resulting in that spatial objects in the created orthoimages cannot be located in their true positions. This paper presents our research and experimental results of true orthoimage generation in extremely tall urban areas using lidar and multi-view large-scale aerial images. Lidar data are used for the extraction of an urban digital surface model (DSM), further for the extraction of a digital building model (DBM) and a digital terrain model (DTM). Data structure and a data model for managing urban spatial objects, such as buildings and bridges, are developed. The photogrammetric geometry is used for the detection of occluded and shadowed areas in true orthoimage generation. For the occluded and shadowed areas, lost information is compensated from a conjugate area in adjacent images, for which a new mosaicking method, which automatically chooses the 'best' imagery and automatically optimizes the seam line, has been developed. Experimental results from central Denver, Colorado and Lower Manhattan, New York City demonstrated that the proposed true orthoimage generation scheme in this paper is capable of truly orthorectifying the relief displacement in aerial images and significantly reducing occlusion and shadow defects.     

Change detection of buildings from satellite imagery and lidar data

Geospatial objects change over time and this necessitates periodic updating of the cartography that represents them. Currently, this updating is done manually, by interpreting aerial photographs, but this is an expensive and time-consuming process. While several kinds of geospatial objects are recognized, this article focuses on buildings. Specifically, we propose a novel automatic approach for detecting buildings that uses satellite imagery and laser scanner data as a tool for updating buildings for a vector geospatial database. We apply the support vector machine (SVM) classification algorithm to a joint satellite and laser data set for the extraction of buildings. SVM training is automatically carried out from the vector geospatial database. For visualization purposes, the changes are presented using a variation of the traffic-light map. The different colours assist human operators in performing the final cartographic updating. Most of the important changes were detected by the proposed method. The method not only detects changes, but also identifies inaccuracies in the cartography of the vector database. Small houses and low buildings surrounded by high trees present significant problems with regard to automatic detection compared to large houses and taller buildings. In addition to visual evaluation, this study was checked for completeness and correctness using numerical evaluation and receiver operating characteristic curves. The high values obtained for these parameters confirmed the efficacy of the method.     

多特征融合的高分辨率遥感影像建筑物分级提取

目的 针对高分辨率遥感影像普遍存在的同谱异物和同物异谱问题,提出一种综合利用光谱、形状、空间上下文和纹理特征的建筑物分级提取方法。方法 该方法基于单幅高分辨率遥感影像,首先利用多尺度多方向梯度算子构造的建筑物指数和形状特征提取部分分割完整的矩形建筑物目标;然后由多方向线性结构元素和形态学膨胀运算确定投票矩阵,从而获取光照方向,并利用光照方向和阴影特征对已提取建筑物进行筛选,剔除非建筑物对象,完成建筑物初提取;最后借助初提取建筑物对象的纹理特征向量建立概率模型,取得像素级建筑物提取结果,将该结果与影像分割相结合实现建筑物提取。结果 选取两幅高分辨率遥感影像进行实验,在建筑物初提取实验中,将本文方法与邻域总变分法和Sobel算子进行对比,实验结果表明,本文方法适用性强,为后提取提供的建筑物样本可靠性更高。在建筑物提取实验中,采用查准率、查全率和F₁分数3个指标进行定量分析,与形态学建筑物指数结合形态学阴影指数算法、邻域总变分结合混合高斯模型和贝叶斯判决算法相比,各项精度指标均得到显著提升,其中查准率提高了2.90个百分点,查全率提高了12.49个百分点,F₁分数则提升了8.84。结论 本文提出的建筑物分级提取方法具备一定抗干扰能力,且提取准确性高,适用性强。     

基于对象的最优尺度建筑物信息提取方法

针对基于像素分析方法不适用于高分辨率影像信息提取的问题,提出一种基于对象的图像分析方法来进行城市建筑信息提取。采用多分辨率图像分割方法得到图像对象,提出非监督的最优尺度判定方法解决单尺度分割造成的欠分割和过分割问题。在对象分类提取过程中,结合LiDAR数据的地形表面高程信息和光谱信息对建筑物进行提取,并利用尺寸、空间位置等信息进行误分类修正。实验区域共提取出18个建筑目标,结果表明所提出的方法有效可行。     

Extraction of landslide-related factors from ASTER imagery and its application to landslide susceptibility mapping

The aim of this study is to extract landslide-related factors from remote-sensing data, such as Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) satellite imagery, and to examine their applicability to landslide susceptibility near Boun, Korea, using a geographic information system (GIS). Landslide was mapped from interpretation of aerial photographs and field surveying. Factors that influence landslide occurrence were extracted from ASTER imagery. The slope, aspect and curvature were calculated from the digital elevation model (DEM) with 25.77 m root mean square error (RMSE), which was derived from ASTER imagery. Lineaments, land-cover and normalized difference vegetation index (NDVI) layers were also estimated from ASTER imagery. Landslide-susceptible areas were analysed and mapped using the occurrence factors by a frequency ratio and logistic regression model. Validation results were 84.78% in frequency ratio and 84.20% in logistic regression prediction accuracy for the susceptibility map with respect to ground-truth data.     

Extracting buildings from and regularizing boundaries in airborne lidar data using connected operators

The location of building boundary is a crucial prerequisite for geographical condition monitoring, urban management, and building reconstruction. This paper presents a framework that employs a series of algorithms to automatically extract building footprints from airborne (light detection and ranging (lidar)) data and image. Connected operators are utilized to extract building regions from lidar data, which would not produce new contours nor change their position and have very good contour-preservation properties. First, the building candidate regions are separated from lidar-derived digital surface model (DSM) based on a new method proposed within this paper using connected operators, and trees are removed based on the normalized difference vegetation index (NDVI) value of image. Then, building boundaries are identified and building boundary lines are traced by ‘sleeve’ line simplification method. Finally, the principal directions of buildings are used to regularize the directions of building boundary lines. International Society for Photogrammetry and Remote Sensing (ISPRS) data sets in Vaihingen whose point spacing is about 0.4 m from urbanized areas were employed to test the proposed framework, and three test areas were selected. A quantitative analysis showed that the method proposed within this paper was effective and the average offset values of simple and complex building boundaries were 0.2–0.4 m and 0.3–0.6 m, respectively.     

The Comparison of Urban Tree Crown Extraction based on Airborne LiDARElevation Difference and High Resolution Imagery

With the support of airborne Light Detection and Ranging (LiDAR) data and high spatial resolution aerial imagery,this paper presents an individual tree extraction method that takes the region of urban as the study area.The elevation difference model originated from LiDAR data was used to extract regions of interest including trees. Then,masking was applied to the high spatial resolution aerial imagery to get the same regions. Besides,image segmentations,based on the marked watershed algorithm,were processed on the high spatial resolution aerial imagery and the elevation difference model separately to extract individual tree crowns. Finally,we took a visual interpretation to delineate tree crowns manually and this result was regarded as the reference crowns map. The extraction accuracies were assessed by comparing the spatial relationships of the reference crowns and the automated delineated tree crowns based on the elevation difference model and the high resolution imagery. The results show that the LiDAR data is developed to improve the efficiency of obtaining forest region that the canopy height model include 85.25% forest information. In addition,the tree crowns extraction accuracy based on the high resolution aerial imagery is 57.14%,while another extraction accuracy based on the elevation difference model is 42.47%,which indicated that the marked watershed algorithm proposed in this paper is effective and the high resolution imagery is better than the elevation difference model to extract tree crowns.     

Comparison of impervious surface area and normalized difference vegetation index as indicators of surface urban heat island effects in Landsat imagery

This paper compares the normalized difference vegetation index (NDVI) and percent impervious surface as indicators of surface urban heat island effects in Landsat imagery by investigating the relationships between the land surface temperature (LST), percent impervious surface area (%ISA), and the NDVI. Landsat Thematic Mapper (TM) and Enhanced Thematic Mapper Plus (ETM+) data were used to estimate the LST from four different seasons for the Twin Cities, Minnesota, metropolitan area. A map of percent impervious surface with a standard error of 7.95% was generated using a normalized spectral mixture analysis of July 2002 Landsat TM imagery. Our analysis indicates there is a strong linear relationship between LST and percent impervious surface for all seasons, whereas the relationship between LST and NDVI is much less strong and varies by season. This result suggests percent impervious surface provides a complementary metric to the traditionally applied NDVI for analyzing LST quantitatively over the seasons for surface urban heat island studies using thermal infrared remote sensing in an urbanized environment.     

A simple method for detection and counting of oil palm trees using high-resolution multispectral satellite imagery

In the past, oil palm density has been determined by manually counting trees every year in oil palm plantations. The measurement of density provides important data related to palm productivity, fertilizer needed, weed control costs in a circle around each tree, labourers needed, and needs for other activities. Manual counting requires many workers and has potential problems related to accuracy. Remote sensing provides a potential approach for counting oil palm trees. The main objective of this study is to build a robust and user-friendly method that will allow oil palm managers to count oil palm trees using a remote sensing technique. The oil palm trees analysed in this study have different ages and densities. QuickBird imagery was applied with the six pansharpening methods and was compared with panchromatic QuickBird imagery. The black and white imagery from a false colour composite of pansharpening imagery was processed in three ways: (1) oil palm tree detection, (2) delineation of the oil palm area using the red band, and (3) counting oil palm trees and accuracy assessment. For oil palm detection, we used several filters that contained a Sobel edge detector; texture analysis co-occurrence; and dilate, erode, high-pass, and opening filters. The results of this study improved upon the accuracy of several previous research studies that had an accuracy of about 90–95%. The results in this study show (1) modified intensity-hue-saturation (IHS) resolution merge is suitable for 16-year-old oil palm trees and have rather high density with 100% accuracy; (2) colour normalized (Brovey) is suitable for 21-year-old oil palm trees and have low density with 99.5% accuracy; (3) subtractive resolution merge is suitable for 15- and 18-year-old oil palm trees and have a rather high density with 99.8% accuracy; (4) PC spectral sharpening with 99.3% accuracy is suitable for 10-year-old oil palm trees and have low density; and (5) for all study object conditions, colour normalized (Brovey) and wavelet resolution merge are two pansharpening methods that are suitable for oil palm tree extraction and counting with 98.9% and 98.4% accuracy, respectively.     

Single tree detection in very high resolution remote sensing data

Tree detection is a major focus in the field of (semi-) automatic extraction of forest information from very high resolution remote sensing data. Many existing tree crown delineation algorithms require a set of seed pixels to start the process of image segmentation. In this study, different methods of obtaining seed pixels (semi-) automatically from orthophotos and digital surface models derived from stereo digital camera imagery are tested and compared. The UltracamD digital camera provides images with a stereo overlap of about 90% and this paper presents a new method of DSM generation based on multiple stereo pairs. The evaluation of the DSMs shows that by using a multiple image approach, also referred to as block-based approach, the quality is significantly increased: the mean difference between the estimated values and 356 measured upper layer tree heights is only 0.77 m with a standard deviation of 2.39 m.

In terms of seed generation, the morph algorithm (2d) used in this paper detected 64% of the trees visible in the aerial photos with an error margin of around 25% both for commission and omission in a dense natural forest. The orthophoto-based local maximum approach generally yielded lower accuracies and more multiple hits than the morph algorithm. 3d seed generation from the block based model returned about 70% correct hits for the upper tree canopy layer. All evaluations are performed based on field measurements and visual aerial photo interpretation. Furthermore, the dependence of successful tree detection on the dominance of a tree within the stand is analyzed. As expected, suppressed trees are more likely to be omitted. The segmentation proved to be useful, as the automatically generated segments had a similar number of correct hits as achieved by visual interpretation, with the only drawback being a higher error of commission.     

Object-based larch tree-crown delineation using high-resolution satellite imagery

Olgan larch is a traditional construction material used for the renovation of historical timber-frame buildings in China. However, acquiring the necessary large-sized larch trees from old-growth forests has become a challenge in China because of the rare and inaccessible distribution of these trees. In recent years, remote sensing imagery has provided a more effective alternative for delineating tree crowns automatically with high accuracy. In this study, an object-based method for delineating old-growth larch tree crowns using Geoeye-1 imagery is developed. Tree crown delineation results are tested and evaluated by field data. In addition, the correlation between delineated tree-crown and basal areas are quantitatively validated to ensure that the developed method can be applied for estimating the distribution of old-growth larch trees. Results demonstrate that the developed object-based larch tree-crown delineation method is reliable, thus providing a new technique for detecting old-growth larch tree resources in Northeastern China.     

Downscaling MODIS-derived water maps with high-precision topographic data in a shallow lake

ABSTRACT

Remotely sensed imagery is the most efficient and widely used data source to monitor the water area changes. However, a trade-off always exists between temporal resolution and spatial resolution for satellite images. Taking the southern Dongting Lake as an example, this study was conducted to develop a method of downscaling the Moderate Resolution Imaging Spectroradiometer (MODIS)-derived coarse spatial resolution water maps in shallow lakes with high-precision digital elevation model. The main principle of the method is to identify and adjust the horizontal location errors of the waterlines extracted from coarse-resolution data by analysing and modifying the elevation leaps using finer-scale topography information. Moving average filter was used to smooth the errors of waterlines caused by the geometric inaccuracies and classification uncertainties of the coarse data. The optimal local window size of the moving average filter was selected automatically using an exponential decay function model and a curvature algorithm for each pixel in the waterlines. In reference to Landsat Thematic Mapper data, the accuracy of the downscaling result is distinctly higher than that of the original MODIS normalized difference water index-derived water maps. The presented method is proved to be an effective tool for acquiring water maps of shallow lake with high spatio-temporal resolution using coarse- or moderate-resolution satellite imagery and high-precision topographic data.