WorldView-2 Images Research Articles

Landsat-8, Advanced Spaceborne Thermal Emission and Reflection Radiometer, and WorldView-3 Multispectral Satellite Imagery for Prospecting Copper-Gold Mineralization in the Northeastern Inglefield Mobile Belt (IMB), Northwest Greenland

Several regions in the High Arctic still lingered poorly explored for a variety of mineralization types because of harsh climate environments and remoteness. Inglefield Land is an ice-free region in northwest Greenland that contains copper-gold mineralization associated with hydrothermal alteration mineral assemblages. In this study, Landsat-8, Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), and WorldView-3 multispectral remote sensing data were used for hydrothermal alteration mapping and mineral prospecting in the Inglefield Land at regional, local, and district scales. Directed principal components analysis (DPCA) technique was applied to map iron oxide/hydroxide, Al/Fe-OH, Mg-Fe-OH minerals, silicification (Si-OH), and SiO2 mineral groups using specialized band ratios of the multispectral datasets. For extracting reference spectra directly from the Landsat-8, ASTER, and WorldView-3 (WV-3) images to generate fraction images of end-member minerals, the automated spectral hourglass (ASH) approach was implemented. Linear spectral unmixing (LSU) algorithm was thereafter used to produce a mineral map of fractional images. Furthermore, adaptive coherence estimator (ACE) algorithm was applied to visible and near-infrared and shortwave infrared (VINR + SWIR) bands of ASTER using laboratory reflectance spectra extracted from the USGS spectral library for verifying the presence of mineral spectral signatures. Results indicate that the boundaries between the Franklinian sedimentary successions and the Etah metamorphic and meta-igneous complex, the orthogneiss in the northeastern part of the Cu-Au mineralization belt adjacent to Dallas Bugt, and the southern part of the Cu-Au mineralization belt nearby Marshall Bugt show high content of iron oxides/hydroxides and Si-OH/SiO2 mineral groups, which warrant high potential for Cu-Au prospecting. A high spatial distribution of hematite/jarosite, chalcedony/opal, and chlorite/epidote/biotite were identified with the documented Cu-Au occurrences in central and southwestern sectors of the Cu-Au mineralization belt. The calculation of confusion matrix and Kappa Coefficient proved appropriate overall accuracy and good rate of agreement for alteration mineral mapping. This investigation accomplished the application of multispectral/multi-sensor satellite imagery as a valuable and economical tool for reconnaissance stages of systematic mineral exploration projects in remote and inaccessible metallogenic provinces around the world, particularly in the High Arctic regions.

Random forest classification of Callicarpa nudiflora from WorldView-3 imagery based on optimized feature space

Taking the Xiushui township of Baisha county in Hainan province as the research area,the random forest algorithm with obvious advantages in feature selection and classification extraction was used to extract the information of the Callicarpa nudiflora planting in the study area. Firstly,four kinds of different characteristic variables were generated based on World View-3 data,including spectral features,principal component features,vegetation index and texture features. Secondly,the spatial distribution of the C. nudiflora in the study area was extracted by remote sensing by random forest classification algorithm. Finally,the feature space of the random forest classification algorithm was optimized based on the feature importance to obtain the best random forest classification results,and this result is compared with the classification result of the random forest algorithm of the unoptimized feature space. The results showed that:①The overall accuracy of the C. nudiflora extracted by World View-3 image was 89. 97%,and the Kappa coefficient was 0. 84,which indicates that the random forest algorithm had higher classification accuracy and better applicability in Hainan C. nudiflora recognition.② The overall accuracy of extracting C. nudiflora with the dimension reduction feature was 90. 4,and the Kappa coefficient was 0. 85,which indicates that the random forest algorithm can effectively select features. At the same time as the feature variable data mining,the precision of the information extraction of the C. nudiflora was still guaranteed,and the operation efficiency was improved. This study provides a new idea,method and technical means for information extraction of cultivated medicinal plant resources in terms of feature selection and method selection.

A Spatial Analysis of Unplanned and Mushrooming Dumpsites and Environmental Governance in Nairobi County

Unplanned and mushrooming dumpsites in Nairobi County are growing at an alarming rate regardless of the socio- economic status resulting in unclean, ugly outlook of the city. The creation of these unplanned dumpsites underscores the governance gaps that exist in the solid waste management (SWM) in Nairobi County. In a period of fourteen years, the unplanned dumpsites have an exponential growth that necessitates intervention in the mechanism of handling solid waste in Nairobi County. Environmental governance in solid waste management is becoming a key pillar in the proper waste management strategies and dominates the development discourse. Most Nairobi County residents are not conscious of proper and well-maintained waste management systems. The purpose of this study was to explicate the spatial analysis of the importance of environmental governance in Solid Waste Management in Nairobi County on why unplanned dumpsites are on the increase. The study area was the two hundred and seventy seven unplanned dumpsites distributed across Nairobi County. Sampling was done through spatial analysis, the data sources were mainly from Quick Bird images of 2003 and 2007, and WorldView-2 images of 2013 and 2017. Other secondary data used were Nairobi County boundary shape-file, Kenya rivers shape-file, Kenya roads shape-file and Kenya settlements shape-file, they were all obtained from Survey of Kenya databases. This paper provides a platform for other researchers to conduct further investigation on the importance of environmental governance to promote sustainable Solid Waste Management in Nairobi County with a view of reducing the unplanned and mushrooming dumpsites. Keywords: Solid Waste Management, Environmental governance, unplanned dumpsites, Nairobi County, Spatial analysis, Solid Waste Management DOI : 10.7176/JEES/9-10-17 Publication date :October 31 st 2019

Improved Piecewise Linear Transformation for Precise Warping of Very-High-Resolution Remote Sensing Images

A large number of evenly distributed conjugate points (CPs) in entirely overlapping regions of the images are required to achieve successful co-registration between very-high-resolution (VHR) remote sensing images. The CPs are then used to construct a non-linear transformation model that locally warps a sensed image to a reference image’s coordinates. Piecewise linear (PL) transformation is largely exploited for warping VHR images because of its superior performance as compared to the other methods. The PL transformation constructs triangular regions on a sensed image from the CPs by applying the Delaunay algorithm, after which the corresponding triangular regions in a reference image are constructed using the same CPs on the image. Each corresponding region in the sensed image is then locally warped to the regions of the reference image through an affine transformation estimated from the CPs on the triangle vertices. The warping performance of the PL transformation shows reliable results, particularly in regions inside the triangles, i.e., within the convex hulls. However, the regions outside the triangles, which are warped when the extrapolated boundary planes are extended using CPs located close to the regions, incur severe geometric distortion. In this study, we propose an effective approach that focuses on the improvement of the warping performance of the PL transformation over the external area of the triangles. Accordingly, the proposed improved piecewise linear (IPL) transformation uses additional pseudo-CPs intentionally extracted from positions on the boundary of the sensed image. The corresponding pseudo-CPs on the reference image are determined by estimating the affine transformation from CPs located close to the pseudo-CPs. The latter are simultaneously used with the former to construct the triangular regions, which are enlarged accordingly. Experiments on both simulated and real datasets, constructed from Worldview-3 and Kompsat-3A satellite images, were conducted to validate the effectiveness of the proposed IPL transformation. That transformation was shown to outperform the existing linear/non-linear transformation models such as an affine, third and fourth polynomials, local weighted mean, and PL. Moreover, we demonstrated that the IPL transformation improved the warping performance over the PL transformation outside the triangular regions by increasing the correlation coefficient values from 0.259 to 0.304, 0.603 to 0.657, and 0.180 to 0.338 in the first, second, and third real datasets, respectively.

Comparison of classification methods for urban green space extraction using very high resolution worldview-3 imagery

Urban green space (UGS) plays a vital role in maintaining the ecological balance of a city and in ensuring healthy living of the city inhabitants. It is generally suggested that one-third of the city should be covered by green and to ensure this, the city administrators must have an accurate map of the existing UGS. Such a map would be useful to visualize the distribution of the existing green cover and also to find out the areas that can possibly be converted to UGS. Reported studies on UGS mapping have mostly used medium and high resolution images such as Landsat-TM, ETM+, Sentinel-2A, IKONOS, etc. However, studies on the use of very high resolution images for UGS extraction are very limited. The present study is a first attempt in utilizing the very high resolution Worldview-3 image for UGS extraction. Performance of different classification methods such as unsupervised, supervised, object based and normalized difference vegetation index (NDVI) were compared using the pan sharpened Worldview-3 image covering part of New Delhi in India. It was found that the unsupervised classification followed by manual recoding method showed superior performance with overall accuracy (OA) of 99% and κ coefficient of 0.98. Also, the OA achieved in the present study is the highest when compared to other reported studies on UGS extraction. The map of UGS revealed that almost 40% of the study area is covered by green which is more than the recommended value of 33% (one-third). In order to check the universality of the unsupervised classification approach in extracting UGS, Worldview-3 image covering Rio in Brazil was tested. It was found that an OA of 98% and κ coefficient of 0.95 were obtained which clearly indicate that the proposed approach would work very well in extracting UGS from any Worldview-3 imagery.

A Comparison of WorldView-2 and Landsat 8 Images for the Classification of Forests Affected by Bark Beetle Outbreaks Using a Support Vector Machine and a Neural Network: A Case Study in the Sumava Mountains

The objective of this paper is to assess WorldView-2 (WV2) and Landsat OLI (L8) images in the detection of bark beetle outbreaks in the Sumava National Park. WV2 and L8 images were used for the classification of forests infected by bark beetle outbreaks using a Support Vector Machine (SVM) and a Neural Network (NN). After evaluating all the available results, the SVM can be considered the best method used in this study. This classifier achieved the highest overall accuracy and Kappa index for both classified images. In the cases of WV2 and L8, total overall accuracies of 86% and 71% and Kappa indices of 0.84 and 0.66 were achieved with SVM, respectively. The NN algorithm using WV2 also produced very promising results, with over 80% overall accuracy and a Kappa index of 0.79. The methods used in this study may be inspirational for testing other types of satellite data (e.g., Sentinel-2) or other classification algorithms such as the Random Forest Classifier.

Comparison of ATCOR Atmospheric and ELM Linear Empirical Correction Models Applied to WorldView-2 Images

The use of remote sensing images for the characterization of the coastal marine ecosystems requires the prior removal of the atmospheric effects, which can be done in a semi-automated manner, by the use of algorithms based on alternative assumptions contained in the processing tools for different software packages. The main objective of this study was to statistically compare the spectral behavior of the coverages contained in an high-resolution WorldView-2 image atmospherically corrected according to the ATCOR and empirical linear models (ELM), using field spectroradiometry conducted in the insular areas of the archipelago of San Andres and Providence. The ATCOR correction model was applied through the PCI 2015 Geomatics software; regarding the ELM model, the ENVI 5.2 software was used. For the spectral comparison four (4) types of coverage were selected (vegetation, reef formations, beach sand and submerged sandbank), with twenty (20) replicas each, for a total of eighty (80) sampling points distributed in a stratified way in the image. The statistical results showed a linear correlation greater than 0.9 between the reflectance values for each of the bands (Blue, Green, Red and NIR-1) and indicate that both models of the atmospheric correction have a high capacity to eliminate the atmospheric effects present in this type of images. However, there are minor significant differences between the middle quadratic errors in the reflectance values for each band of the corrected images.

An Operational Workflow of Deciduous-Dominated Forest Species Classification: Crown Delineation, Gap Elimination, and Object-Based Classification

Recent advances in remote sensing technology provide sufficient spatial detail to achieve species-level classification over large vegetative ecosystems. In deciduous-dominated forests, however, as tree species diversity and forest structural diversity increase, the frequency of spectral overlap between species also increases and our ability to classify tree species significantly decreases. This study proposes an operational workflow of individual tree-based species classification for a temperate, mixed deciduous forest using three-seasonal WorldView images, involving three steps of individual tree crown (ITC) delineation, non-forest gap elimination, and object-based classification. The process of species classification started with ITC delineation using the spectral angle segmentation algorithm, followed by object-based random forest classifications. A total of 672 trees was located along three triangular transects for training and validation. For single-season images, the late-spring, mid-summer, and early-fall images achieve the overall accuracies of 0.46, 0.42, and 0.35, respectively. Combining the spectral information of the early-spring, mid-summer, and early-fall images increases the overall accuracy of classification to 0.79. However, further adding the late-fall image to separate deciduous and coniferous trees as an extra step was not successful. Compared to traditional four-band (Blue, Green, Red, Near-Infrared) images, the four additional bands of WorldView images (i.e., Coastal, Yellow, Red Edge, and Near-Infrared2) contribute to the species classification greatly (OA: 0.79 vs. 0.53). This study gains insights into the contribution of the additional spectral bands and multi-seasonal images to distinguishing species with seemingly high degrees of spectral overlap.

Mapping Honeysuckle Distribution in Large Forests through Use of High Resolution Satellite Images – Case Study at Great Parks of Hamilton County, Ohio, USA

A non-native invasive shrub, named Amur honeysuckle, has caused various problems to the habitats of native plants and animals. Local governments, private landowners and non-profit groups spend numerous efforts on removing these species from forests. The first step to remove the species is to know their spatial distribution. The major objective of this study is to explore how to quickly map honeysuckle distribution by using high resolution satellite images. The case study is located in Great Parks of Hamilton County (GPHC), Ohio. A World-View2 satellite image acquired on Nov. 9th 2013 was classified into several land use types, including low density honeysuckle area, medium/high density honeysuckle area, grass/crop, bare soil/construction, and water/wet land. Half of the field data that we collected in Nov. 2013 and some high resolution aerial photos (with 4 inch spatial resolution) were used to validate the classification result. The study result demonstrates that the World-View2 image delineates honeysuckle distribution fairly well, in particular, for medium/high density honeysuckle areas. As a pilot program utilizing advanced geospatial analysis, this project will provide important information for understanding the status of wildlife habitats and for implementing site-specific management in parks and nature preserves.

Integration of WorldView-2 and airborne laser scanning data to classify defoliation levels in Quercus ilex L. Dehesas affected by root rot mortality: Management implications

The dehesa ecosystem is severely affected by root rot decline caused by the soilborne pathogenic oomycetes of the genus Phytophthora. Defoliation mapping is particularly challenging in low density ecosystems with high spectral variability due to environmental heterogeneity within relatively short distances. In this study, holm oak defoliation was assessed by classifying two different levels (low and high defoliation trees, with a 30% crown-defoliation threshold) using a combination of multispectral WorldView-2 (WV-2) and Airborne Laser Scanning (ALS) data in an automated Random Forest (RF) modeling approach. Besides the original remote sensing data, a set of vegetation indices (VI) and ALS-derived structural parameters were derived. Seven predictors of defoliation were selected from the ALS metrics, WV-2 bands, and indices datasets: the height percentile of the tree in which occurs the maximum number of returns, mean values in the tree crown of the Chlorophyll index green, Plant Senescence Reflectance Index, Red Edge Normalized Difference Vegetation Index, Double Peak Index, band 3 of the WV-2 image, and bincentil 70 values. The confusion matrix of the external evaluation achieved the highest overall classification accuracy (86.7%) and the highest Kappa index (0.73). Finally, we produced a predicted tree-defoliation map which was used to establish the relationship between Quercus ilex damage levels and soil variables and management practices. High defoliated trees were more abundant in zones of shallow and compacted soil, with lower silt proportion and Ca concentration. Our results offer to forest managers a tool for the rapid and effective assessment of large areas affected by root rot defoliation in their planning of control and restoration measures aimed at reducing the holm oak mortality of holm oaks in management units.

Land-Cover Classification of Coastal Wetlands Using the RF Algorithm for Worldview-2 and Landsat 8 Images

Wetlands are one of the world’s most important ecosystems, playing an important role in regulating climate and protecting the environment. However, human activities have changed the land cover of wetlands, leading to direct destruction of the environment. If wetlands are to be protected, their land cover must be classified and changes to it monitored using remote sensing technology. The random forest (RF) machine learning algorithm, which offers clear advantages (e.g., processing feature data without feature selection and preferable classification result) for high spatial image classification, has been used in many study areas. In this research, to verify the effectiveness of this algorithm for remote sensing image classification of coastal wetlands, two types of spatial resolution images of the Linhong Estuary wetland in Lianyungang—Worldview-2 and Landsat-8 images—were used for land cover classification using the RF method. To demonstrate the preferable classification accuracy of the RF algorithm, the support vector machine (SVM) and k-nearest neighbor (k-NN) methods were also used to classify the same area of land cover for comparison with the results of RF classification. The study results showed that (1) the overall accuracy of the RF method reached 91.86%, higher than the SVM and k-NN methods by 4.68% and 4.72%, respectively, for Worldview-2 images; (2) at the same time, the classification accuracies of RF, SVM, and k-NN were 86.61%, 79.96%, and 77.23%, respectively, for Landsat-8 images; (3) for some land cover types having only a small number of samples, the RF algorithm also achieved better classification results using Worldview-2 and Landsat-8 images, and (4) the addition texture features could improve the classification accuracy of the RF method when using Worldview-2 images. Research indicated that high-resolution remote sensing images are more suitable for small-scale land cover classification image and that the RF algorithm can provide better classification accuracy and is more suitable for coastal wetland classification than the SVM and k-NN algorithms are.

Context-Enabled Extraction of Large-Scale Urban Functional Zones from Very-High-Resolution Images: A Multiscale Segmentation Approach

Urban functional-zone (UFZ) analysis has been widely used in many applications, including urban environment evaluation, and urban planning and management. How to extract UFZs’ spatial units which delineates UFZs’ boundaries is fundamental to urban applications, but it is still unresolved. In this study, an automatic, context-enabled multiscale image segmentation method is proposed for extracting spatial units of UFZs from very-high-resolution satellite images. First, a window independent context feature is calculated to measure context information in the form of geographic nearest-neighbor distance from a pixel to different image classes. Second, a scale-adaptive approach is proposed to determine appropriate scales for each UFZ in terms of its context information and generate the initial UFZs. Finally, the graph cuts algorithm is improved to optimize the initial UFZs. Two datasets including WorldView-2 image in Beijing and GaoFen-2 image in Nanchang are used to evaluate the proposed method. The results indicate that the proposed method can generate better results from very-high-resolution satellite images than widely used approaches like image tiles and road blocks in representing UFZs. In addition, the proposed method outperforms existing methods in both segmentation quality and running time. Therefore, the proposed method appears to be promising and practical for segmenting large-scale UFZs.

Determining Effective Meter-Scale Image Data and Spectral Vegetation Indices for Tropical Forest Tree Species Differentiation

We evaluate several high-resolution remote sensing data products derived from WorldView-3 satellite imagery to determine which product or product grouping would be most applicable for identifying and mapping tropical tree species. The study site, La Selva Biological Station in Costa Rica, provides the infrastructure to facilitate collection of proper ground-truth data for image validation. An objective statistical analysis demonstrated that the WorldView-3 imagery, after applying a series of spectral and illumination image corrections, was able to accurately identify selected tree species. Specifically, this study defines the image bands and image-derived spectral vegetation indices that are the most effective for tree species discrimination. We show that corrected absolute reflectivity values from the green, red, red edge, and near-infrared are good differentiators of tropical tree species crowns. We also evaluate 14 possible multiband vegetation indices and show that two new indices developed here, using WorldView-3 image bands, have the highest discriminatory power for tropical tree species. A combination of both individual band information and vegetation indices would significantly improve image-based classification of tropical forests.

Simple and effective cloud- and shadow-detection algorithms for Landsat and Worldview images

Many conventional cloud- and shadow-detection algorithms require meta-data such as sun angle and date of image collection. Moreover, detection results can vary a lot in actual images. We present simple and effective algorithms that do not require meta-data for detecting clouds and shadows in Landsat and Worldview images. Comparison with existing state-of-the-art algorithms, including a deep learning-based algorithm as well as a commercial algorithm, using actual satellite images, shows that the simple algorithms have comparable or even better performance than existing algorithms.

Comparing Human Versus Machine-Driven Cadastral Boundary Feature Extraction

The objective to fast-track the mapping and registration of large numbers of unrecorded land rights globally has led to the experimental application of Artificial Intelligence in the domain of land administration, and specifically the application of automated visual cognition techniques for cadastral mapping tasks. In this research, we applied and compared the ability of rule-based systems within Object-Based Image Analysis (OBIA), as opposed to human analysis, to extract visible cadastral boundaries from very high-resolution World View-2 images, in both rural and urban settings. From our experiments, machine-based techniques were able to automatically delineate a good proportion of rural parcels with explicit polygons where the correctness of the automatically extracted boundaries was 47.4% against 74.24% for humans and the completeness of 45% for the machine compared to 70.4% for humans. On the contrary, in the urban area, automatic results were counterintuitive: even though urban plots and buildings are clearly marked with visible features such as fences, roads and tacitly perceptible to eyes, automation resulted in geometrically and topologically poorly structured data. Thus, these could neither be geometrically compared with human digitisation, nor actual cadastral data from the field. The results of this study provide an updated snapshot with regards to the performance of contemporary machine-driven feature extraction techniques compared to conventional manual digitising. In our methodology, using an iterative approach of segmentation and classification, we demonstrated how to overcome the weaknesses of having undesirable segments due to intra-parcel and inter-parcel variability, when using segmentation approaches for cadastral feature delineation. We also demonstrated how we can easily implement a geometric comparison framework within the Esri’s ArcGIS software environment and firmly believe the developed methodology can be reproduced.

Vegetation patterns in a South American coastal wetland using high-resolution imagery

ABSTRACT The aim of this study was to identify and characterize the main plant communities in a temperate coastal wetland using high-resolution imagery. We produced a map of Samborombón Bay at 1:25,000 scale using a WorldView-2 image. An Object-based Image Analysis approach was chosen, and an unsupervised classification algorithm was applied. Overall classification accuracy was 81%, and the Kappa index was 78.1%. Six land cover types were mapped including four main natural monospecific plant communities. The lower intertidal area was dominated by mudflats without vegetation and stands of Bolboschoenus maritimus. The middle intertidal area was dominated by Sarcocornia ambigua, while, in the higher intertidal area, Sporobolus densiflorus and grasslands with Cortaderia selloana prevailed. We found four spatial patterns at a landscape scale, based on the presence and spatial distribution of the natural plant communities. This map represents a valuable tool for future studies on wetland environmental indicators.

От идеи до образа: об одном механизме эволюции научной идеи

<p>В статье анализируется проблема визуализированного образа как сознательно созданного зрительного представления определенной идеи. Он отличается от модели, чертежа, диаграммы, эскиза тем, что воспринимается человеком как объективно существующий предмет или процесс (обладает свойством интенциональности). В отличие от ощущения визуализированный образ является изображением, обладает структурным соответствием объекту. В научном познании визуализированные образы выполняют две функции: а) они способствуют пониманию теоретического знания; б) являются важным средством трансляции научных идей во вненаучные сферы (прежде всего в область образования).</p><p>Предлагается следующий алгоритм трансформации научной идеи в визуализированный образ: 1) создание теоретических конструктов исследуемого пространства при помощи абстрагирования или идеализации; 2) установление связи между теоретическими конструктами, создание теоретической схемы; 3) изображение теоретической схемы при помощи средств графической наглядности; 4) формирование визуализированного образа. В случае включения визуализированного образа в научную картину мира он приобретает статус мировоззренческого образа.</p>

The Comparison of Fusion Methods for HSRRSI Considering the Effectiveness of Land Cover (Features) Object Recognition Based on Deep Learning

The efficient and accurate application of deep learning in the remote sensing field largely depends on the pre-processing technology of remote sensing images. Particularly, image fusion is the essential way to achieve the complementarity of the panchromatic band and multispectral bands in high spatial resolution remote sensing images. In this paper, we not only pay attention to the visual effect of fused images, but also focus on the subsequent application effectiveness of information extraction and feature recognition based on fused images. Based on the WorldView-3 images of Tongzhou District of Beijing, we apply the fusion results to conduct the experiments of object recognition of typical urban features based on deep learning. Furthermore, we perform a quantitative analysis for the existing pixel-based mainstream fusion methods of IHS (Intensity-Hue Saturation), PCS (Principal Component Substitution), GS (Gram Schmidt), ELS (Ehlers), HPF (High-Pass Filtering), and HCS (Hyper spherical Color Space) from the perspectives of spectrum, geometric features, and recognition accuracy. The results show that there are apparent differences in visual effect and quantitative index among different fusion methods, and the PCS fusion method has the most satisfying comprehensive effectiveness in the object recognition of land cover (features) based on deep learning.

Application of feature selection methods and machine learning algorithms for saltmarsh biomass estimation using Worldview-2 imagery

Assessing large scale plant productivity of coastal marshes is essential to understand the resilience of these systems to climate change. Two machine learning approaches, random forest (RF) and support vector machine (SVM) regression were tested to estimate biomass of a common saltmarshes species, salt couch grass (Sporobolus virginicus). Reflectance and vegetation indices derived from 8 bands of Worldview-2 multispectral data were used for four experiments to develop the biomass model. These four experiments were, Experiment-1: 8 bands of Worldview-2 image, Experiment-2: Possible combination of all bands of Worldview-2 for Normalized Difference Vegetation Index (NDVI) type vegetation indices, Experiment-3: Combination of bands and vegetation indices, Experiment-4: Selected variables derived from experiment-3 using variable selection methods. The main objectives of this study are (i) to recommend an affordable low cost data source to predict biomass of a common saltmarshes species, (ii) to suggest a variable selection method suitable for multispectral data, (iii) to assess the performance of RF and SVM for the biomass prediction model. Cross-validation of parameter optimizations for SVM showed that optimized parameter of ɛ-SVR failed to provide a reliable prediction. Hence, ν-SVR was used for the SVM model. Among the different variable selection methods, recursive feature elimination (RFE) selected a minimum number of variables (only 4) with an RMSE of 0.211 (kg/m2). Experiment-4 (only selected bands) provided the best results for both of the machine learning regression methods, RF (R 2= 0.72, RMSE= 0.166 kg/m2) and SVR (R 2= 0.66, RMSE = 0.200 kg/m2) to predict biomass. When a 10-fold cross validation of the RF model was compared with a 10-fold cross validation of SVR, a significant difference (p = <0.0001) was observed for RMSE. One to one comparisons of actual to predicted biomass showed that RF underestimates the high biomass values, whereas SVR overestimates the values; this suggests a need for further investigation and refinement.

Unsupervised Deep Change Vector Analysis for Multiple-Change Detection in VHR Images

Change detection (CD) in multitemporal images is an important application of remote sensing. Recent technological evolution provided very high spatial resolution (VHR) multitemporal optical satellite images showing high spatial correlation among pixels and requiring an effective modeling of spatial context to accurately capture change information. Here, we propose a novel unsupervised context-sensitive framework—deep change vector analysis (DCVA)—for CD in multitemporal VHR images that exploit convolutional neural network (CNN) features. To have an unsupervised system, DCVA starts from a suboptimal pretrained multilayered CNN for obtaining deep features that can model spatial relationship among neighboring pixels and thus complex objects. An automatic feature selection strategy is employed layerwise to select features emphasizing both high and low prior probability change information. Selected features from multiple layers are combined into a deep feature hypervector providing a multiscale scene representation. The use of the same pretrained CNN for semantic segmentation of single images enables us to obtain coherent multitemporal deep feature hypervectors that can be compared pixelwise to obtain deep change vectors that also model spatial context information. Deep change vectors are analyzed based on their magnitude to identify changed pixels. Then, deep change vectors corresponding to identified changed pixels are binarized to obtain a compressed binary deep change vectors that preserve information about the direction (kind) of change. Changed pixels are analyzed for multiple CD based on the binary features, thus implicitly using the spatial information. Experimental results on multitemporal data sets of Worldview-2, Pleiades, and Quickbird images confirm the effectiveness of the proposed method.