Very High Resolution Images Research Articles

An Object-Based Change Detection Approach Using Uncertainty Analysis for VHR Images

This paper proposes an object-based approach to supervised change detection using uncertainty analysis for very high resolution (VHR) images. First, two temporal images are combined into one image by band stacking. Then, on the one hand, the stacked image is segmented by the statistical region merging (SRM) to generate segmentation maps; on the other hand, the stacked image is classified by the support vector machine (SVM) to produce a pixel-wise change detection map. Finally, the uncertainty analysis for segmented objects is implemented to integrate the segmentation map and pixel-wise change map at the appropriate scale and generate the final change map. Experiments were carried out with SPOT 5 and QuickBird data sets to evaluate the effectiveness of proposed approach. The results indicate that the proposed approach often generates more accurate change detection maps compared with some methods and reduces the effects of classification and segment scale on the change detection accuracy. The proposed method supplies an effective approach for the supervised change detection for VHR images.

A study of a Gaussian mixture model for urban land-cover mapping based on VHR remote sensing imagery

ABSTRACTThis article proposes a Gaussian-mixture-model (GMM)-based method with optimal Gaussian components to address the high intra-class spectral variability in urban land-cover mapping using remote sensing images with very high resolution (VHR). GMMs can simulate and approximate any data distribution provided the optimal Gaussian components can be found. Through improving the model parameters in view of the characteristic of VHR remote sensing images, the parameter space of GMM is optimized significantly, and the model can find the optimal Gaussian components that are suitable for remote sensing images with different resolutions. Experimental results of Wuhan urban area using two images with different resolutions have demonstrated the efficiency and effectiveness of the model. The optimized GMM-based method performs at least comparably or superior to the state-of-the-art classifiers such as support vector machines (SVMs), characterizes man-made land-cover types better than conventional methods, fuses spectral and textural features of VHR image properly, and meanwhile has lower computational complexity.

An Approach to Fine Coregistration Between Very High Resolution Multispectral Images Based on Registration Noise Distribution

Even after applying effective coregistration methods, multitemporal images are likely to show a residual misalignment, which is referred to as registration noise (RN). This is because coregistration methods from the literature cannot fully handle the local dissimilarities induced by differences in the acquisition conditions (e.g., the stability of the acquisition platform, the off-nadir angle of the sensor, the structure of the considered scene, etc.). This paper addresses the problem of reducing such a residual misalignment by proposing a fine automatic coregistration approach for very high resolution (VHR) multispectral images. The proposed method takes advantage of the properties of the residual misalignment itself. To this end, RN is first extracted in the change vector analysis (CVA) polar domain according to the behaviors of the specific multitemporal images considered. Then, a local analysis of RN pixels (i.e., those showing residual misalignment) is conducted for automatically extracting control points (CPs) and matching them according to their estimated displacement. Matched CPs are used for generating a deformation map by interpolation. Finally, one VHR image is warped to the coordinates of the other through a deformation map. Experiments carried out on simulated and real multitemporal VHR images confirm the effectiveness of the proposed approach.

Estimating the Numbers and the Areas of Collapsed Buildings by Combining VHR Images, Statistics and Survey Data: a Case Study of the Lushan Earthquake in China

Accurately obtaining the structures and damage types of buildings in earthquake stricken areas is fundamental to supporting rescue forces and estimating economic losses and casualties. As the stricken areas are often much larger than the areas covered by very high resolution (VHR) images, the information obtained from VHR images cannot satisfy practical needs. This study developed a method for estimating the structures and types of damaged buildings by combining VHR images, statistics and ground survey data. First, the rates of damaged buildings with different structures and damage types were manually interpreted from VHR images covering a small part of the stricken area, and further corrected by ground survey data. Second, the corrected rates were reallocated to the seismic intensity zones. Third, the rates in the seismic intensity zones and the statistical data were combined to estimate the numbers and areas of damaged buildings in villages, towns and counties. The presented method was applied to estimate the damages caused by the Lushan earthquake in China. The results indicated that our method can efficiently estimate the amount of the damages and complement existing work on only automatic extracting damaged buildings from VHR images.

A scene change detection framework for multi-temporal very high resolution remote sensing images

The technology of computer vision and image processing is attracting more and more attentions in recent years, and has been applied in many research areas like remote sensing image analysis. Change detection with multi-temporal remote sensing images is very important for the dynamic analysis of landscape variations. The abundant spatial information offered by very high resolution (VHR) images makes it possible to identify the semantic classes of image scenes. Compared with the traditional approaches, scene change detection can provide a new point of view for the semantic interpretation of land-use transitions. In this paper, for the first time, we explore a scene change detection framework for VHR images, with a bag-of-visual-words (BOVW) model and classification-based methods. Image scenes are represented by a word frequency with three kinds of multi-temporal learned dictionary, i.e., the separate dictionary, the stacked dictionary, and the union dictionary. Three features (multispectral raw pixel; mean and standard deviation; and SIFT) and their combinations were tested in scene change detection. Post-classification and compound classification were evaluated for their performances in the “from–to” change results. Two multi-temporal scene datasets were used to quantitatively evaluate the proposed scene change detection approach. The results indicate that the proposed scene change detection framework can obtain a satisfactory accuracy and can effectively analyze land-use changes, from a semantic point of view.

A Direction-Guided Ant Colony Optimization Method for Extraction of Urban Road Information From Very-High-Resolution Images

Typical object-based classification methods only take image object properties as criteria to classify roads, leaving the associated edge information unused. These methods often lead to fragmented road areas and inconsistent road widths and smoothness. Meanwhile, very-high-resolution (VHR) images contain a large amount of edge information and different types of geographic objects, thus, it is challenging to extract roads by typical edge-based extraction or grouping methods. In this study, a globally optimized method is developed to integrate both object and edge features to extract urban road information from VHR images. This novel method extends ant colony optimization (ACO) through deploying and moving ants (artificial agents) along roads with the guidance of comprehensive object and edge information. As ants spread pheromone along their paths, roads are recognized based on aggregated pheromone levels. A set of experiments on VHR images showed that our method significantly outperforms object-based classification methods with not only improved road extraction quality but also enhanced stability when applied to large and complex images.

Measuring the Effectiveness of Various Features for Thematic Information Extraction From Very High Resolution Remote Sensing Imagery

Generally, some object-based features are more relevant to a thematic class than other features. These strongly relevant features, termed as class-specific features, would significantly contribute to thematic information extraction for very high resolution (VHR) images. However, many existing feature selection methods have been designed to select a good feature subset for all classes, rather than an independent feature subset for the thematic class. The latter might better meet the requirement of thematic information extraction than the former. In addition, the lack of quantitative evaluation of the contribution of the selected features to thematic classes also weakens our understandability of these features. To address the problems, class-specific feature selection methods are developed to measure the effectiveness of features for extracting thematic information from VHR images. First, the one-versus-all scheme is combined with traditional feature selection methods, such as ReliefF and LeastC. Also, one-versus-one scheme is utilized for alleviating the negative impact of a class imbalance problem arising from the one-versus-all scheme. Then, the relative contributions of features to thematic classes are obtained by the class-specific feature selection methods to describe the effectiveness of features for thematic information extraction. Finally, the class-specific feature selection methods are compared with the original methods on three different VHR image data sets by the nearest neighbor and support vector machine. Experimental results show that the class-specific feature selection methods outperform the corresponding conventional methods, and the one-versus-one scheme surpasses one-versus-all scheme. Additionally, many features are evaluated by the class-specific feature selection methods, to provide end users advice on effectiveness of the features.

Orthorectification and Pan-Sharpening of WorldView-2 Satellite Imagery to Produce High Resolution Coloured Ortho-Photos

In the last decade VHR (Very High Resolution) images from satellite, because of the reduced dimensions of pixel (less than 1 meter) and the availability in different acquisition bands (4 or more), have had major dffusion in many application fields of remote sensing. They can be used also to produce high resolution coloured ortho-photos, but adequate levels of positional accuracy as well as small pixel dimensions are necessary. The aim of this paper is to demonstrate that WorldView-2 (WV-2) images satisfy totally these requirements if firstly submitted to high accurate rectification and Pan-Sharpening processes. Using Rational Polynomial Functions (RPFs), original dataset can be better overlapped to cartographic maps at medium or great scale; multispectral images (cell size: 2 m) can be resampled to meet geometric resolution of pan one (cell size: 0.5 m), so detailed and attendible RGB composition results. Applications are carried out on one sample of WV-2 imagery concerning a scene within the Province of Caserta (Italy) that includes vegetated as well as urban areas. Finally RGB composition with pixel dimensions of 0.5 m, positional accuracy less than 1 meter and likely colors are achieved, confirming the possibility to use this type of images for coloured ortho-photos at scale 1:5.000 at least.

Using random walker for knowledge transfer in classifying multi-temporal VHR images

Acquiring land cover types from very high resolution (VHR) images is of great significance to many applications and has been intensively studied for many years. The difficulties in image classification and the high frequencies of remote sensing image acquisition make it urgent to develop efficient knowledge transfer approaches for understanding multi-temporal VHR images. This letter proposed a knowledge transfer approach that uses the label information of the existing VHR images to classify multi-temporal images. The approach was implemented in three steps: object-based change detection, knowledge transfer of label information, and random walker (RW) classification. The proposed approach was tested by two datasets with each having two temporal images acquired on the same geographical areas. The experimental results showed that the proposed approach outperformed the support vector machine (SVM) algorithm in classifying multi-temporal images and can reduce the influence of spectral confusions on image classification.

Spatiotemporal inferences for use in building detection using series of very-high-resolution space-borne stereo images

Automatic building detection from very-high-resolution (VHR) satellite images is a difficult task. The detection accuracy is usually limited by spectral ambiguities and the uncertainties of the available height information. Feature extraction and training sampling collection for supervised methods are other sources of uncertainty. Most widely used VHR sensors have shorter revisit cycles (IKONOS/GeoEye-1/2, 3 days; WorldView 1/2, 1.1 days) due to large off-nadir viewing angles and hence are able to perform consistent acquisition of mono or stereo images. In this article, we investigate the possibility of using high-temporal stereo VHR images to enhance remote-sensing image interpretation under the context of building detection. Digital surface models, which contain the height information, are generated for each date using semi-global matching. Pre-classification is performed combining the height and spectral information to obtain an initial building probability map. With a reference land cover map available for one date, the training samples of the other dates are automatically derived using a rule-based validating procedure. A spatiotemporal inference filter is developed considering the spectral, spatial, and temporal aspects to enhance the building probability maps. This aims at homogenizing the building probability values of spectrally similar pixels in the spatial domain and geometrically similar pixels in the temporal domain, while being robust to the silhouette of the images and geometric discrepancies of the multitemporal data. The effectiveness and robustness of the proposed method are evaluated by performing three experiments on six stereo pairs of the same region over a time period of five years (2006–2011). The area under curve (AUC) of the receiver operating characteristic and kappa statistic (κ) are employed to assess the results. These experiments show that spatiotemporal inference filtering largely improves the accuracy of the building probability map (average AUC = 0.95) while facilitating building extraction in snow-covered images. The resulting building probability maps can be further used for other applications (e.g. building footprint updating).

Exploring GIS knowledge to improve building extraction and change detection from VHR imagery in urban areas

Existing studies for building extraction in very high resolution (VHR) images consider little prior knowledge, and thus they have limited accuracies and can only apply to a small proportion of buildings. Combining VHR and GIS data to extract buildings is a feasible way to overcome the drawbacks above. However, the inaccurate positions of GIS data and time changes between the two data make it difficult to fuse them. This study aims at presenting the methods to resolve the position and time inconsistencies between the two data for extracting buildings and finding the changes. The methods begin with line extraction from VHR images. For extracting buildings, a two-level graph with four steps was presented to fuse GIS contours and the extracted line segments from VHR images, including choosing initial building edges, clustering the chosen edges, generating the building hypotheses and scoring building hypotheses. The time inconsistencies (i.e. the changes between the buildings in two data sources) were first identified by a scoring function. Then, the detected inconsistent buildings were further verified by the z-significance test from a statistical perspective. Experimental assessments performed on 10 typical urban regions indicated that the proposed methods are highly robust and convincing.

Mapping of Agricultural Crops from Single High-Resolution Multispectral Images—Data-Driven Smoothing vs. Parcel-Based Smoothing

Mapping agricultural crops is an important application of remote sensing. However, in many cases it is based either on hyperspectral imagery or on multitemporal coverage, both of which are difficult to scale up to large-scale deployment at high spatial resolution. In the present paper, we evaluate the possibility of crop classification based on single images from very high-resolution (VHR) satellite sensors. The main objective of this work is to expose performance difference between state-of-the-art parcel-based smoothing and purely data-driven conditional random field (CRF) smoothing, which is yet unknown. To fulfill this objective, we perform extensive tests with four different classification methods (Support Vector Machines, Random Forest, Gaussian Mixtures, and Maximum Likelihood) to compute the pixel-wise data term; and we also test two different definitions of the pairwise smoothness term. We have performed a detailed evaluation on different multispectral VHR images (Ikonos, QuickBird, Kompsat-2). The main finding of this study is that pairwise CRF smoothing comes close to the state-of-the-art parcel-based method that requires parcel boundaries (average difference ≈ 2.5%). Our results indicate that a single multispectral (R, G, B, NIR) image is enough to reach satisfactory classification accuracy for six crop classes (corn, pasture, rice, sugar beet, wheat, and tomato) in Mediterranean climate. Overall, it appears that crop mapping using only one-shot VHR imagery taken at the right time may be a viable alternative, especially since high-resolution multitemporal or hyperspectral coverage as well as parcel boundaries are in practice often not available.

Development of Line-of-Sight Digital Surface Model for Co-Registering Off-Nadir VHR Satellite Imagery With Elevation Data

Co-registration of very-high-resolution (VHR) images with elevation data is extremely important for many remote sensing applications due to the complementary properties of these two data types. However, this type of multidata source registration has many associated challenges. For instance, although VHR satellite images are usually acquired off-nadir, the integration of off-nadir images with digital surface models (DSMs) for the purpose of urban mapping has been rarely seen in research publications. This is due to the relief displacement of the elevated objects, which causes a problematic misregistration between the perspective off-nadir images and the corresponding orthographic DSMs. Therefore, the co-registration of such datasets is almost impossible unless a true orthorectification process is executed. However, true orthoimages are expensive, time consuming, and difficult to achieve. Thus, this paper proposes a registration method based on developing a line-of-sight DSM solution to effectively register elevation data with off-nadir VHR images. The method utilizes the relevant sensor model in two phases: deriving DSM from stereo images and reprojecting the DSM back to one of the stereo images to generate a line-of-sight DSM for accurate co-registration. To demonstrate the applicability of the proposed method and evaluate the effect of the misregistration, a building detection procedure is implemented. The proposed method is found to be feasible, inexpensive, and of subpixel accuracy. Additionally, it improves the overall accuracy of detecting buildings by almost 12% relative to that when the conventional two-dimensional (2-D) registration technique is used solely due to the elimination of the misregistration effect.

COMPARISON OF UNSUPERVISED VEGETATION CLASSIFICATION METHODS FROM VHR IMAGES AFTER SHADOWS REMOVAL BY INNOVATIVE ALGORITHMS

Abstract. The recognition of vegetation by the analysis of very high resolution (VHR) aerial images provides meaningful information about environmental features; nevertheless, VHR images frequently contain shadows that generate significant problems for the classification of the image components and for the extraction of the needed information. The aim of this research is to classify, from VHR aerial images, vegetation involved in the balance process of the environmental biochemical cycle, and to discriminate it with respect to urban and agricultural features. Three classification algorithms have been experimented in order to better recognize vegetation, and compared to NDVI index; unfortunately all these methods are conditioned by the presence of shadows on the images. Literature presents several algorithms to detect and remove shadows in the scene: most of them are based on the RGB to HSI transformations. In this work some of them have been implemented and compared with one based on RGB bands. Successively, in order to remove shadows and restore brightness on the images, some innovative algorithms, based on Procrustes theory, have been implemented and applied. Among these, we evaluate the capability of the so called “not-centered oblique Procrustes” and “anisotropic Procrustes” methods to efficiently restore brightness with respect to a linear correlation correction based on the Cholesky decomposition. Some experimental results obtained by different classification methods after shadows removal carried out with the innovative algorithms are presented and discussed.

Extraction of earthquake-induced collapsed buildings using very high-resolution imagery and airborne lidar data

Extraction of urban building damage caused by earthquake disasters, from very-high-resolution (VHR) satellite imagery and related geospatial data, has been widely studied in the past decade. In this study, a multi-stage collapsed building detection method, using bi-temporal (pre- and post-earthquake) VHR images and post-earthquake airborne light detection and ranging (lidar) data, is proposed. Ground objects that are intact and significantly different from collapsed buildings, such as intact buildings, pavements, shadows, and vegetation, were first extracted using the post-event VHR image and lidar data and masked out. Collapsed buildings were then extracted by classifying the combined bi-temporal VHR images and texture images of the remaining area using a one-class classifier, the One-Class Support Vector Machine (OCSVM). A post-processing procedure was adopted to refine the obtained result. The proposed method was quantitatively evaluated and compared to two existing methods in Port au Prince, Haiti, which was heavily hit by an earthquake in January 2010. In the two comparative methods, data for the whole study area were directly used. In the first method, collapsed buildings were extracted directly using the OCSVM, while in the second method, buildings and pavements were removed from the extraction result of the first method. The results showed that the proposed method significantly outperformed the existing methods, with increases of 21% and 40%, respectively, in the kappa coefficient. The proposed method provides a fast and reliable method to detect collapsed urban buildings caused by earthquake disasters, and could also be applied to other study areas using similar data combinations.

Testing Texture of VHR Panchromatic Data as a Feature of Land Cover Classification

While it is well-known that texture can be used to classify very high resolution (VHR) data, the limits of its applicability have not been unequivocally specified. This study examines whether it is possible to divide satellite images into two classes associated with “low” and “high” texture values in the initial stage of processing VHR images. This approach can be effectively used in object-oriented classification. Based on the panchromatic channel of KOMPSAT-2 images from five areas of Europe, datasets with down-sampled pixel resolutions of 1, 2, 4, 8, and 16 m were prepared. These images were processed using different texture analysis techniques in order to discriminate between basic land cover classes. Results were assessed using the normalized feature space distance expressed by the Jeffries–Matusita distance. The best results were observed for images with the highest resolution processed by the Laplacian filter. Our research shows that a classification approach based on the idea of “low” and “high” textures can be effectively applied to panchromatic data with a resolution of 8 m or higher.

Object-Oriented Semisupervised Classification of VHR Images by Combining MedLDA and a Bilateral Filter

A Bayesian hierarchical model is presented to classify very high resolution (VHR) images in a semisupervised manner, in which both a maximum entropy discrimination latent Dirichlet allocation (MedLDA) and a bilateral filter are combined into a novel application framework. The primary contribution of this paper is to nullify the disadvantages of traditional probabilistic topic models on pixel-level supervised information and to achieve the effective classification of VHR remote sensing images. This framework consists of the following two iterative steps. In the training stage, the model utilizes the central labeled pixel and its neighborhood, as a squared labeled image object, to train the classifiers. In the classification stage, each central unlabeled pixel with its neighborhood, as an unlabeled object, is classified as a user-provided geoobject class label with the maximum posterior probability. Gibbs sampling is adopted for model inference. The experimental results demonstrate that the proposed method outperforms two classical SVM-based supervised classification methods and probabilistic-topic-models-based classification methods.

Multiscale Region-Level VHR Image Change Detection via Sparse Change Descriptor and Robust Discriminative Dictionary Learning.

Very high resolution (VHR) image change detection is challenging due to the low discriminative ability of change feature and the difficulty of change decision in utilizing the multilevel contextual information. Most change feature extraction techniques put emphasis on the change degree description (i.e., in what degree the changes have happened), while they ignore the change pattern description (i.e., how the changes changed), which is of equal importance in characterizing the change signatures. Moreover, the simultaneous consideration of the classification robust to the registration noise and the multiscale region-consistent fusion is often neglected in change decision. To overcome such drawbacks, in this paper, a novel VHR image change detection method is proposed based on sparse change descriptor and robust discriminative dictionary learning. Sparse change descriptor combines the change degree component and the change pattern component, which are encoded by the sparse representation error and the morphological profile feature, respectively. Robust change decision is conducted by multiscale region-consistent fusion, which is implemented by the superpixel-level cosparse representation with robust discriminative dictionary and the conditional random field model. Experimental results confirm the effectiveness of the proposed change detection technique.

A new segmentation method for very high resolution imagery using spectral and morphological information

Image segmentation is a key and prerequisite step for object-based analysis of very high resolution (VHR) imagery. Most existing image segmentation methods use either spectral or spatial information of an image alone. A novel image segmentation method for VHR multispectral images using combined spectral and morphological information is proposed in this paper. The method can be summarized as follows. First, a morphological derivative profile is calculated from an original multispectral image and combined with the spectral bands to quantify spectral-morphological characteristics of a pixel, which are considered as a criterion of homogeneity of neighboring pixels. Image segmentation is then conducted using a seeded region-growing procedure, which is based on the seed points automatically generated from the gradient image and dynamically added and the similarity between a seed pixel and its neighboring pixels in terms of spectral-morphological characteristics. The obtained segmentation result is further refined by a region merging procedure to generate a final segmentation result. The proposed method is evaluated using three VHR images of urban and suburban areas and compared with two existing segmentation methods, in terms of visual inspection, quantitative evaluation and indirect evaluation. Experimental results demonstrate that the joint use of spectral and morphological information outperformed the use of morphological information alone. Furthermore, the proposed image segmentation method performed better than existing methods. The proposed image segmentation method is well applicable to the segmentation of VHR imagery over urban and suburban areas.

Aboveground biomass mapping of African forest mosaics using canopy texture analysis: toward a regional approach.

In the context of the reduction of greenhouse gas emissions caused by deforestation and forest degradation (the REDD+ program), optical very high resolution (VHR) satellite images provide an opportunity to characterize forest canopy structure and to quantify aboveground biomass (AGB) at less expense than methods based on airborne remote sensing data. Among the methods for processing these VHR images, Fourier textural ordination (FOTO) presents a good method to detect forest canopy structural heterogeneity and therefore to predict AGB variations. Notably, the method does not saturate at intermediate AGB values as do pixelwise processing of available space borne optical and radar signals. However, a regional-scale application requires overcoming two difficulties: (1) instrumental effects due to variations in sun–scene–sensor geometry or sensor-specific responses that preclude the use of wide arrays of images acquired under heterogeneous conditions and (2) forest structural diversity including monodominant or open canopy forests, which are of particular importance in Central Africa. In this study, we demonstrate the feasibility of a rigorous regional study of canopy texture by harmonizing FOTO indices of images acquired from two different sensors (Geoeye-1 and QuickBird-2) and different sun–scene–sensor geometries and by calibrating a piecewise biomass inversion model using 26 inventory plots (1 ha) sampled across very heterogeneous forest types. A good agreement was found between observed and predicted AGB (residual standard error [RSE] = 15%; R2 = 0.85; P < 0.001) across a wide range of AGB levels from 26 Mg/ha to 460 Mg/ha, and was confirmed by cross validation. A high-resolution biomass map (100-m pixels) was produced for a 400-km2 area, and predictions obtained from both imagery sources were consistent with each other (r = 0.86; slope = 1.03; intercept = 12.01 Mg/ha). These results highlight the horizontal structure of forest canopy as a powerful descriptor of the entire forest stand structure and heterogeneity. In particular, we show that quantitative metrics resulting from such textural analysis offer new opportunities to characterize the spatial and temporal variation of the structure of dense forests and may complement the toolbox used by tropical forest ecologists, managers or REDD+ national monitoring, reporting and verification bodies.