Cidália Costa Fonte

Usability of VGI for validation of land cover maps

Volunteered Geographic Information (VGI) represents a growing source of potentially valuable data for many applications, including land cover map validation. It is still an emerging field and many different approaches can be used to take value from VGI, but also many pros and cons are related to its use. Therefore, since it is timely to get an overview of the subject, the aim of this article is to review the use of VGI as reference data for land cover map validation. The main platforms and types of VGI that are used and that are potentially useful are analysed. Since quality is a fundamental issue in map validation, the quality procedures used by the platforms that collect VGI to increase and control data quality are reviewed and a framework for addressing VGI quality assessment is proposed. A review of cases where VGI was used as an additional data source to assist in map validation is made, as well as cases where only VGI was used, indicating the procedures used to assess VGI quality and fitness for use. A discussion and some conclusions are drawn on best practices, future potential and the challenges of the use of VGI for land cover map validation.

Areas of fuzzy geographical entities

Fuzzy Geographical Entities (FGEs) refer in this paper to geographical entities with fuzzy spatial extent. The use of FGEs in geographical information systems requires the existence of operators capable of processing them. In this paper, our contribution to that field focuses on the computation of areas. Two methods are considered, one crisp due to Rosenfeld (1984), which has limited applicability, and the other fuzzy, which is a new approach. The new fuzzy area operator gives more information about the possible values of the area and enables the fuzziness in the spatial extent of the entity to be propagated to the area. Crisp and fuzzy areas have different meanings, and the use of one or the other depends not only on the purpose of the computation but also on the semantics of the membership functions. When the FGEs are represented by normal fuzzy sets, the fuzzy area operator generates fuzzy numbers, and therefore arithmetic operations can be performed with them using fuzzy arithmetic. However, we show that care must be taken with the use of the fuzzy arithmetic operators because, in some situations, the usual operators should not be applied. Properties of the Rosenfeld and fuzzy area operators are analysed, establishing a parallel with properties of the areas of crisp sets.

A method to incorporate uncertainty in the classification of remote sensing images

The aim of this paper is to investigate if the incorporation of the uncertainty associated with the classification of surface elements into the classification of landscape units (LUs) increases the results accuracy. To this end, a hybrid classification method is developed, including uncertainty information in the classification of very high spatial resolution multi-spectral satellite images, to obtain a map of LUs. The developed classification methodology includes the following steps: (1) a pixel-based hard classification with a probabilistic Bayesian classifier; (2) computation of the posterior probabilities and quantification of the classification uncertainty using an uncertainty measure; (3) image segmentation and (4) object classification based on decision rules. The classification of the resulting objects into LUs is performed considering a set of decision rules that incorporate the pixel-based classification uncertainty. The proposed methodology was tested on the classification of an IKONOS satellite image. The accuracy of the classification was computed using an error matrix. The comparison between the results obtained with the proposed approach and those obtained without considering the classification uncertainty revealed a 12% increase in the overall accuracy. This shows that the information about uncertainty can be valuable when making decisions and can actually increase the accuracy of the classification results.

Repurposing a deep learning network to filter and classify volunteered photographs for land cover and land use characterization

Abstract This paper extends recent research into the usefulness of volunteered photos for land cover extraction, and investigates whether this usefulness can be automatically assessed by an easily accessible, off-the-shelf neural network pre-trained on a variety of scene characteristics. Geo-tagged photographs are sometimes presented to volunteers as part of a game which requires them to extract relevant facts about land use. The challenge is to select the most relevant photographs in order to most efficiently extract the useful information while maintaining the engagement and interests of volunteers. By repurposing an existing network which had been trained on an extensive library of potentially relevant features, we can quickly carry out initial assessments of the general value of this approach, pick out especially salient features, and identify focus areas for future neural network training and development. We compare two approaches to extract land cover information from the network: a simple post hoc weighting approach accessible to non-technical audiences and a more complex decision tree approach that involves training on domain-specific features of interest. Both approaches had reasonable success in characterizing human influence within a scene when identifying the land use types (as classified by Urban Atlas) present within a buffer around the photograph’s location. This work identifies important limitations and opportunities for using volunteered photographs as follows: (1) the false precision of a photograph’s location is less useful for identifying on-the-spot land cover than the information it can give on neighbouring combinations of land cover; (2) ground-acquired photographs, interpreted by a neural network, can supplement plan view imagery by identifying features which will never be discernible from above; (3) when dealing with contexts where there are very few exemplars of particular classes, an independent a posteriori weighting of existing scene attributes and categories can buffer against over-specificity.

Modelling the Fuzzy Spatial Extent of Geographical Entities

In several situations the spatial extent of geographical entities is uncertain or fuzzy. In such cases the entities may be represented using fuzzy sets through the construction of the herein called “fuzzy geographical entities”. Four sources of fuzziness are identified in the process of constructing geographical entities characterized by predefined attributes through the classification of a tessellation. For each, a method to compute the membership grades to the fuzzy geographical entity is proposed, based on the appropriate semantic interpretation of the grades of membership. The interpretations used are the likelihood view of membership grades, the random set view and the similarity view. A practical example is presented for each case.

Temporal Evolution of Sunspot Areas and Estimation of Related Plasma Flows

The increased amount of information provided by ongoing missions such as the Solar Dynamics Observatory (SDO) represents a great challenge for the understanding of basic questions such as the internal structure of sunspots and how they evolve with time. Here, we contribute with the exploitation of new data, to provide a better understanding of the separate growth and decay of sunspots, umbra, and penumbra. Using fuzzy sets to compute separately the areas of sunspot umbra and penumbra, the growth and decay rates for active regions NOAA 11117, NOAA 11428, NOAA 11429, and NOAA 11430 are computed from the analysis of intensitygrams obtained by the Helioseismic and Magnetic Imager onboard SDO. A simplified numerical model is proposed for the decay phase, whereby an empirical irrotational and uniformly convergent horizontal velocity field interacting with an axially symmetric and height-invariant magnetic field reproduces the large-scale features of the much more complex convection observed inside sunspots.

The Role of Citizen Science in Earth Observation

Citizen Science (CS) and crowdsourcing are two potentially valuable sources of data for Earth Observation (EO), which have yet to be fully exploited. Research in this area has increased rapidly during the last two decades, and there are now many examples of CS projects that could provide valuable calibration and validation data for EO, yet are not integrated into operational monitoring systems. A special issue on the role of CS in EO has revealed continued trends in applications, covering a diverse set of fields from disaster response to environmental monitoring (land cover, forests, biodiversity and phenology). These papers touch upon many key challenges of CS including data quality and citizen engagement as well as the added value of CS including lower costs, higher temporal frequency and use of the data for calibration and validation of remotely-sensed imagery. Although still in the early stages of development, CS for EO clearly has a promising role to play in the future.

Using OpenStreetMap to Create Land Use and Land Cover Maps

OpenStreetMap (OSM) is a bottom up community-driven initiative to create a global map of the world. Yet the application of OSM to land use and land cover (LULC) mapping is still largely unexploited due to problems with inconsistencies in the data and harmonization of LULC nomenclatures with OSM. This chapter outlines an automated methodology for creating LULC maps using the nomenclature of two European LULC products: the Urban Atlas (UA) and CORINE Land Cover (CLC). The method is applied to two regions in London and Paris. The results show that LULC maps with a level of detail similar to UA can be obtained for the urban regions, but that OSM has limitations for conversion into the more detailed non-urban classes of the CLC nomenclature. Future work will concentrate on developing additional rules to improve the accuracy of the transformation and building an online system for processing the data.

Assessing the applicability of OpenStreetMap data to assist the validation of land use/land cover maps

ABSTRACT The validation of land use/land cover (LULC) maps is usually performed using a reference database consisting of a sample of points or regions to which the ‘real’ class is assigned. This assignment is usually performed by specialists using photointerpretation (PI) of high-resolution imagery and/or field visits, which are time consuming and expensive processes. The aim of this article is to assess if the data available in the collaborative project OpenStreetMap (OSM) may be used as a source of data to assist the creation of these reference databases, reducing the time spent and costs associated with their generation. For this aim, two case studies were used, where the validation of the Global Monitoring for Environment and Security Urban Atlas (UA) was performed. The used methodology requires the harmonization of the data available in OSM with the UA nomenclature, and the subsequent creation of a LULC map from the OSM data. This map was then compared to UA to assess the similarity of the regions mapped in both. To test the usefulness of OSM data to assess the accuracy of UA, a sample of points was created and two reference databases generated, one assigning the data extracted automatically from OSM to the points where these data were available, and PI for the remaining points, and the other using only PI. The accuracy assessment of UA for the two case studies was then made building confusion matrixes and computing accuracy indicators. The results showed that for the two study areas, only low percentages of points had to be photo interpreted in the first reference database (respectively, 12% and 2% for the two study areas), decreasing the work load considerably. The results obtained with both reference databases are comparable for level 1 classes. For level 2 classes, worse results were obtained for some classes, showing that the OSM data used are not enough to create reliable reference data.

Evaluation of soft possibilistic classifications with non-specificity uncertainty measures

The aim of this paper was to investigate the usefulness of non-specificity uncertainty measures to evaluate soft classifications of remote sensing images. In particular, we analysed whether these measures could be used to identify the difficulties found by the classifier and to estimate the classification accuracy. Two non-specificity uncertainty measures were considered, the non-specificity measure (NSp) and the U-uncertainty measure, and their behaviour was analysed to evaluate which is the most appropriate for this application. To overcome the fact that these two measures have different ranges, a normalized version (Un) of the U-uncertainty measure was used. Both measures were applied to evaluate the uncertainty of a soft classification of a very high spatial resolution multispectral satellite image, performed with an object-oriented image analysis based on a fuzzy classification. The classification accuracy was evaluated using an error matrix and the users and producers accuracies were computed. Two uncertainty indexes are proposed for each measure, and the correlation between the information given by them and the users and producers accuracies was determined to assess the relationship and compatibility of both sources of information. The results show that there is a positive correlation between the information given by the uncertainty and accuracy indexes, but mainly between the uncertainty indexes and the users accuracy, where the correlation achieved 77%. This study shows that uncertainty indexes may be used, along with the possibility distributions, as indicators of the classification performance, and may therefore be very useful tools.