The contribution of multitemporal information from multispectral satellite images for automatic land cover classification at the national scale

Thesis submitted to the Instituto Superior de Estatística e Gestão de Informação da Universidade Nova de Lisboa in partial fulfillment of the requirements for the Degree of Doctor of Philosophy in Information Management – Geographic Information Systems

Land cover classification of Lago Grande de Curuai floodplain (Amazon, Brazil) using multi-sensor and image fusion techniques

Given the limitations of different types of remote sensing images, automated land-cover classifications of the Amazon várzea may yield poor accuracy indexes. One way to improve accuracy is through the combination of images from different sensors, by either image fusion or multi-sensor classifications. Therefore, the objective of this study was to determine which classification method is more efficient in improving land cover classification accuracies for the Amazon várzea and similar wetland environments - (a) synthetically fused optical and SAR images or (b) multi-sensor classification of paired SAR and optical images. Land cover classifications based on images from a single sensor (Landsat TM or Radarsat-2) are compared with multi-sensor and image fusion classifications. Object-based image analyses (OBIA) and the J.48 data-mining algorithm were used for automated classification, and classification accuracies were assessed using the kappa index of agreement and the recently proposed allocation and quantity disagreement measures. Overall, optical-based classifications had better accuracy than SAR-based classifications. Once both datasets were combined using the multi-sensor approach, there was a 2% decrease in allocation disagreement, as the method was able to overcome part of the limitations present in both images. Accuracy decreased when image fusion methods were used, however. We therefore concluded that the multi-sensor classification method is more appropriate for classifying land cover in the Amazon várzea.

Land-cover classification from airborne LIDAR data fused with aerial optical images

Airborne LIght Detection And Ranging (LIDAR) provides accurate height information for objects on the earth, which makes LIDAR become more and more popular in terrain and land surveying. In particular, LIDAR data offer vital and significant features for land-cover classification which is an important task in many application domains. In this paper, an unsupervised approach based on an improved fuzzy Markov random field (FMRF) model is developed, by which the LIDAR data, its co-registered images acquired by optical sensors, i.e. aerial color image and near infrared image, and other derived features are fused effectively to improve the ability of the LIDAR system for the accurate land-cover classification. In the proposed FMRF model-based approach, the spatial contextual information is applied by modeling the image as a Markov random field (MRF), with which the fuzzy logic is introduced simultaneously to reduce the errors caused by the hard classification. Moreover, a Lagrange-Multiplier (LM) algorithm is employed to calculate a maximum A posteriori (MAP) estimate for the classification. The experimental results have proved that fusing the height data and optical images is particularly suited for the land-cover classification. The proposed approach works very well for the classification from airborne LIDAR data fused with its coregistered optical images and the average accuracy is improved to 88.9%.

An effective approach for land-cover classification from airborne lidar fused with co-registered data

Airborne lidar provides accurate height information of objects on the earth and has been recognized as a reliable and accurate surveying tool in many applications. In particular, lidar data offer vital and significant features for urban land-cover classification, which is an important task in urban land-use studies. In this article, we present an effective approach in which lidar data fused with its co-registered images (i.e. aerial colour images containing red, green and blue (RGB) bands and near-infrared (NIR) images) and other derived features are used effectively for accurate urban land-cover classification. The proposed approach begins with an initial classification performed by the Dempster–Shafer theory of evidence with a specifically designed basic probability assignment function. It outputs two results, i.e. the initial classification and pseudo-training samples, which are selected automatically according to the combined probability masses. Second, a support vector machine (SVM)-based probability estimator is adopted to compute the class conditional probability (CCP) for each pixel from the pseudo-training samples. Finally, a Markov random field (MRF) model is established to combine spatial contextual information into the classification. In this stage, the initial classification result and the CCP are exploited. An efficient belief propagation (EBP) algorithm is developed to search for the global minimum-energy solution for the maximum a posteriori (MAP)-MRF framework in which three techniques are developed to speed up the standard belief propagation (BP) algorithm. Lidar and its co-registered data acquired by Toposys Falcon II are used in performance tests. The experimental results prove that fusing the height data and optical images is particularly suited for urban land-cover classification. There is no training sample needed in the proposed approach, and the computational cost is relatively low. An average classification accuracy of 93.63% is achieved.

Toward Satellite-Based Land Cover Classification Through Optimum-Path Forest

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Land cover classification of Lago Grande de Curuai floodplain (Amazon, Brazil) using multi-sensor and image fusion techniques

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

A comparison of pixel and object-based land cover classification. A case study of the Asmara region, Eritrea

In this paper we compare the performance of two image classification paradigms (object- and pixel-based) for creating a land cover map of Asmara, the capital of Eritrea and its surrounding areas using a Landsat ETM+ imagery acquired in January 2000. The image classification methods used were maximum likelihood for the pixel-based approach and Bhattacharyya distance for the object-oriented approach available in, respectively, ArcGIS and SPRING software packages. Advantages and limitations of both approaches are presented and discussed. Classifications outputs were assessed using overall accuracy and Kappa indices. Pixel- and object-based classification methods result in an overall accuracy of 78% and 85%, respectively. The Kappa coefficient for pixel- and object-based approaches was 0.74 and 0.82, respectively. Although pixel-based approach is the most commonly used method, assessment and visual interpretation of the results clearly reveal that the object-oriented approach has advantages for this specific case-study.

A Texture-based land cover classification for the delineation of a shifting cultivation landscape in the Lao PDR using landscape metrics

The delineation of shifting cultivation landscapes using remote sensing in mountainous regions is challenging. On the one hand, there are difficulties related to the distinction of forest and fallow forest classes as occurring in a shifting cultivation landscape in mountainous regions. On the other hand, the dynamic nature of the shifting cultivation system poses problems to the delineation of landscapes where shifting cultivation occurs. We present a two-step approach based on an object-oriented classification of Advanced Land Observing Satellite, Advanced Visible and Near-Infrared Spectrometer (ALOS AVNIR) and Panchromatic Remote-sensing Instrument for Stereo Mapping (ALOS PRISM) data and landscape metrics. When including texture measures in the object-oriented classification, the accuracy of forest and fallow forest classes could be increased substantially. Based on such a classification, landscape metrics in the form of land cover class ratios enabled the identification of crop-fallow rotation characteristics of the shifting cultivation land use practice. By classifying and combining these landscape metrics, shifting cultivation landscapes could be delineated using a single land cover dataset.

The Lena River Delta - Land cover classification of tundra environments based on Landsat 7 ETM+ data and its application for upscaling of methane emissions

The Lena River Delta, situated in Northern Siberia (72.0 - 73.8° N, 122.0 - 129.5° E), is the largest Arctic delta and covers 29,000 km**2. Since natural deltas are characterised by complex geomorphological patterns and various types of ecosystems, high spatial resolution information on the distribution and extent of the delta environments is necessary for a spatial assessment and accurate quantification of biogeochemical processes as drivers for the emission of greenhouse gases from tundra soils. In this study, the first land cover classification for the entire Lena Delta based on Landsat 7 Enhanced Thematic Mapper (ETM+) images was conducted and used for the quantification of methane emissions from the delta ecosystems on the regional scale. The applied supervised minimum distance classification was very effective with the few ancillary data that were available for training site selection. Nine land cover classes of aquatic and terrestrial ecosystems in the wetland dominated (72%) Lena Delta could be defined by this classification approach. The mean daily methane emission of the entire Lena Delta was calculated with 10.35 mg CH4/m**2/d. Taking our multi-scale approach into account we find that the methane source strength of certain tundra wetland types is lower than calculated previously on coarser scales.

Land cover classification of Samoylov Island and Landsat subpixel water cover of Lena River Delta, Siberia, with links to ESRI grid files

Ignoring small-scale heterogeneities in Arctic land cover may bias estimates of water, heat and carbon fluxes in large-scale climate and ecosystem models. We investigated subpixel-scale heterogeneity in CHRIS/PROBA and Landsat-7 ETM+ satellite imagery over ice-wedge polygonal tundra in the Lena Delta of Siberia, and the associated implications for evapotranspiration (ET) estimation. Field measurements were combined with aerial and satellite data to link fine-scale (0.3 m resolution) with coarse-scale (upto 30 m resolution) land cover data. A large portion of the total wet tundra (80%) and water body area (30%) appeared in the form of patches less than 0.1 ha in size, which could not be resolved with satellite data. Wet tundra and small water bodies represented about half of the total ET in summer. Their contribution was reduced to 20% in fall, during which ET rates from dry tundra were highest instead. Inclusion of subpixel-scale water bodies increased the total water surface area of the Lena Delta from 13% to 20%. The actual land/water proportions within each composite satellite pixel was best captured with Landsat data using a statistical downscaling approach, which is recommended for reliable large-scale modelling of water, heat and carbon exchange from permafrost landscapes.

Textural classification of land cover using support vector machines: an empirical comparison with parametric, non parametric and hybrid classifiers in the Bolivian Amazon

Land cover classification is a key research field in remote sensing and land change science as thematic maps derived from remotely sensed data have become the basis for analyzing many socio-ecological issues. However, land cover classification remains a difficult task and it is especially challenging in heterogeneous tropical landscapes where nonetheless such maps are of great importance. The present study aims to establish an efficient classification approach to accurately map all broad land cover classes in a large, heterogeneous tropical area of Bolivia, as a basis for further studies (e.g., land cover-land use change). Specifically, we compare the performance of parametric (maximum likelihood), non-parametric (k-nearest neighbour and four different support vector machines - SVM), and hybrid classifiers, using both hard and soft (fuzzy) accuracy assessments. In addition, we test whether the inclusion of a textural index (homogeneity) in the classifications improves their performance. We classified Landsat imagery for two dates corresponding to dry and wet seasons and found that non-parametric, and particularly SVM classifiers, outperformed both parametric and hybrid classifiers. We also found that the use of the homogeneity index along with reflectance bands significantly increased the overall accuracy of all the classifications, but particularly of SVM algorithms. We observed that improvements in producer’s and user’s accuracies through the inclusion of the homogeneity index were different depending on land cover classes. Earlygrowth/degraded forests, pastures, grasslands and savanna were the classes most improved, especially with the SVM radial basis function and SVM sigmoid classifiers, though with both classifiers all land cover classes were mapped with producer’s and user’s accuracies of around 90%. Our approach seems very well suited to accurately map land cover in tropical regions, thus having the potential to contribute to conservation initiatives, climate change mitigation schemes such as REDD+, and rural development policies.

Domain Adaptation in remote sensing: increasing the portability of land-cover classifiers

Among the types of remote sensing acquisitions, optical images are certainly one of the most widely relied upon data sources for Earth observation. They provide detailed measurements of the electromagnetic radiation reflected or emitted by each pixel in the scene. Through a process termed supervised land-cover classification, this allows to automatically yet accurately distinguish objects at the surface of our planet. In this respect, when producing a land-cover map of the surveyed area, the availability of training examples representative of each thematic class is crucial for the success of the classification procedure. However, in real applications, due to several constraints on the sample collection process, labeled pixels are usually scarce. When analyzing an image for which those key samples are unavailable, a viable solution consists in resorting to the ground truth data of other previously acquired images. This option is attractive but several factors such as atmospheric, ground and acquisition conditions can cause radiometric differences between the images, hindering therefore the transfer of knowledge from one image to another. The goal of this Thesis is to supply remote sensing image analysts with suitable processing techniques to ensure a robust portability of the classification models across different images. The ultimate purpose is to map the land-cover classes over large spatial and temporal extents with minimal ground information. To overcome, or simply quantify, the observed shifts in the statistical distribution of the spectra of the materials, we study four approaches issued from the field of machine learning. First, we propose a strategy to intelligently sample the image of interest to collect the labels only in correspondence of the most useful pixels. This iterative routine is based on a constant evaluation of the pertinence to the new image of the initial training data actually belonging to a different image. Second, an approach to reduce the radiometric differences among the images by projecting the respective pixels in a common new data space is presented. We analyze a kernel-based feature extraction framework suited for such problems, showing that, after this relative normalization, the cross-image generalization abilities of a classifier are highly increased. Third, we test a new data-driven measure of distance between probability distributions to assess the distortions caused by differences in the acquisition geometry affecting series of multi-angle images. Also, we gauge the portability of classification models through the sequences. In both exercises, the efficacy of classic physically- and statistically-based normalization methods is discussed. Finally, we explore a new family of approaches based on sparse representations of the samples to reciprocally convert the data space of two images. The projection function bridging the images allows a synthesis of new pixels with more similar characteristics ultimately facilitating the land-cover mapping across images.

Spatial and temporal analysis of land cover and landscape structure change in Zagros Forests

The Zagros oak forests in Western Iran are critically important to the sustainability of the region. These forests have undergone dramatic declines in recent decades. We evaluated the utility of the non-parametric Random Forest classification algorithm for land cover classification of Zagros landscapes, and selected the best spatial and spectral predictive variables. The algorithm resulted in high overall classification accuracies (>85%) and also equivalent classification accuracies for the datasets from the three different sensors. We evaluated the associations between trends in forest area and structure with trends in socioeconomic and climatic conditions, to identify the most likely driving forces creating deforestation and landscape structure change. We used available socioeconomic (urban and rural population, and rural income), and climatic (mean annual rainfall and mean annual temperature) data for two provinces in northern Zagros. The most correlated driving force of forest area loss was urban population, and climatic variables to a lesser extent. Landscape structure changes were more closely associated with rural population. We examined the effects of scale changes on the results from spatial pattern analysis. We assessed the impacts of eight years of protection in a protected area in northern Zagros at two different scales (both grain and extent). The effects of protection on the amount and structure of forests was scale dependent. We evaluated the nature and magnitude of changes in forest area and structure over the entire Zagros region from 1972 to 2009. We divided the Zagros region in 167 Landscape Units and developed two measures— Deforestation Sensitivity (DS) and Connectivity Sensitivity (CS) — for each landscape unit as the percent of the time steps that forest area and ECA experienced a decrease of greater than 10% in either measure. A considerable loss in forest area and connectivity was detected, but no sudden (nonlinear) changes were detected at the spatial and temporal scale of the study. Connectivity loss occurred more rapidly than forest loss due to the loss of connecting patches. More connectivity was lost in southern Zagros due to climatic differences and different forms of traditional land use.