Automated crop identification tools are of interest to a wide range of applications related to the environment and agriculture including the monitoring of related policies such as the European Common Agriculture Policy. In this context, this work presents a parcel-based crop classification system which leverages on 1D convolutional neural network supervised learning capacity. For the training and evaluation of the model, we employ open and free data: (i) time series of Sentinel-2 optical data selected to cover the crop season of one year, and (ii) a cadastre-derived database providing detailed delineation of parcels. By considering the most dominant crop types and the temporal features of the optical data, the proposed lightweight approach discriminates a considerable number of crops with high accuracy.

Abstract Spatially consistent and up-to-date maps of human settlements are crucial for addressing policies related to urbanization and sustainability, especially in the era of an increasingly urbanized world. The availability of open and free Sentinel-2 data of the Copernicus Earth Observation program offers a new opportunity for wall-to-wall mapping of human settlements at a global scale. This paper presents a deep-learning-based framework for a fully automated extraction of built-up areas at a spatial resolution of 10 m from a global composite of Sentinel-2 imagery. A multi-neuro modeling methodology building on a simple Convolution Neural Networks architecture for pixel-wise image classification of built-up areas is developed. The core features of the proposed model are the image patch of size 5 × 5 pixels adequate for describing built-up areas from Sentinel-2 imagery and the lightweight topology with a total number of 1,448,578 trainable parameters and 4 2D convolutional layers and 2 flattened layers. The deployment of the model on the global Sentinel-2 image composite provides the most detailed and complete map reporting about built-up areas for reference year 2018. The validation of the results with an independent reference dataset of building footprints covering 277 sites across the world establishes the reliability of the built-up layer produced by the proposed framework and the model robustness. The results of this study contribute to cutting-edge research in the field of automated built-up areas mapping from remote sensing data and establish a new reference layer for the analysis of the spatial distribution of human settlements across the rural–urban continuum.

This letter introduces the new European Settlement Map (ESM) production workflow, presents some indicatory results, and discusses the validation process. Unlike the previous ESM versions, the built-up (BU) extraction is realized through supervised learning (not only by means of image filtering and processing techniques) based on textural and morphological features. Input data are the Copernicus very-high-resolution image collection for the reference year 2015 coming from a variety of sensors. The workflow is fully automated, and it does not include any postprocessing. For the first time, a new layer containing nonresidential buildings was derived by using only remote sensing imagery and training data. The produced BU map is delivered at 2-m-pixel resolution (level-1 layer), while the residential/nonresidential layer (level 2) is delivered at 10-m spatial resolution. More than 46 000 scenes were processed, and around 6 million km2of the European territory was mapped. The workflow was executed on the JRC Big Data platform. The validation showed a balanced accuracy of 0.81 and 0.91 for level 1 and 2 layers, respectively, and 0.71 for only the nonresidential layer.

The recent multiplication of open access initiatives to Big Data from Space is giving momentum to the field by widening substantially the spectrum of scientific communities and users, as well as awareness among the public, while offering new benefits at all levels from individual citizens to the whole society. Following a detailed and rigorous review process, 14 articles have been selected out of 48 submissions for this special issue. These are briefly summarized.

This paper presents a processing chain for handling big volume of remotely sensing data for generating wide extent mosaics. More specifically, the data under consideration are level-1 ground range detected Sentinel-1 products with dual polarisation (VV+VH or HH+HV). Two approaches for a) distribution discretization accompanied by false color composition and b) image rendering and mosaicking are proposed. While these two components are necessary constituents of the presented mosaicking workflow, they can operate independently of each other. The design of the processing chain satisfies three objectives: i) contrasting derivative products of the input Sentinel-1 imagery such as the Global Human Settlement Layer, ii) adapting on a high-throughput computing system for fast execution, and iii) allowing potential extensions to more complex applications such as the image classification. Fast processing, process automation, incremental adjustment and information distinction are the main advantages of the proposed method. Elaboration and focus on these features are carried out during the presentation of the results.

The increasing amount of free and open geospatial data of interest to major societal questions calls for the development of innovative data-intensive computing platforms for the efficient and effective extraction of information from these data. This paper proposes a versatile petabyte-scale platform based on commodity hardware and equipped with open-source software for the operating system, the distributed file system, and the task scheduler for batch processing as well as the containerization of user specific applications. Interactive visualization and processing based on deferred processing are also proposed. The versatility of the proposed platform is illustrated with a series of applications together with their performance metrics.

A new built-up presence index based on the density of corners is presented. Corners are detected by multi scale Harris detector based on Differential Morphological Decompositions before being aggregated in a density of corners. Experiments are conducted with several high resolution panchromatic images and the density of corners is shown to be highly correlated to the PanTex built-up index [1].

Edges are crucial descriptors for image analysis and rely mostly on local spectral distances. In this paper, local spectral changes are assimilated to the local statistical dependences which are measured by the local mutual information. This metric is shown to be invariant to unknown bijective transforms which makes it a good candidate for analyzing consistently satellite images which are affected by uncontrolled spectral distortions. Such an edge property is assessed with WorldView-2 multi-angular sequence of images, where spectral distortions arise with far nadir view angles.

Roofless buildings are encountered in case of conflicts and disasters as well as construction sites. A methodology for characterizing and counting roofless buildings in very high resolution optical images is presented. Using morphological transform to extract specific image components, the proposed method spatially aggregates them in a fuzzy logic framework. The result is a map of roofless building membership. This membership map is then used with partial ground truth to estimate the total number of roofless buildings. Finally, the methods are validated with a WorldView panchromatic image, where the membership map and the counting estimate show good accuracies.

Differential area profiles (DAPs) are point-based multiscale descriptors used in pattern analysis and image segmentation. They are defined through sets of size-based connected morphological filters that constitute a joint area opening top-hat and area closing bottom-hat scale-space of the input image. The work presented in this paper explores the properties of this image decomposition through sets of area zones. An area zone defines a single plane of the DAP vector field and contains all the peak components of the input image, whose size is between the zone's attribute extrema. Area zones can be computed efficiently from hierarchical image representation structures, in a way similar to regular attribute filters. Operations on the DAP vector field can then be computed without the need for exporting it first, and an example with the leveling-like convex/concave segmentation scheme is given. This is referred to as the one-pass method and it is demonstrated on the Max-Tree structure. Its computational performance is tested and compared against conventional means for computing differential profiles, relying on iterative application of area openings and closings. Applications making use of the area zone decomposition are demonstrated in problems related to remote sensing and medical image analysis.

Very High Resolution (VHR) satellite images are products which are very useful for assessing humanitarian situations following a disaster or a crisis. When disaster or conflict hits, generally several square kilometers on Earth surface are involved. Having a high resolution and an extended surface, the scenes, from which the information is extracted, become huge in the number of pixels. Built-up structures are the first visible objects from these images which give information about the situation. In this paper, an interactive image mining paradigm is presented for detecting structures of interest in gigapixel images. The proposed approach is tested for refugee tents detection in submetric panchromatic image.

A new unsupervised change detection method for modeling nonlinear temporal dependences based on local information is proposed. A theoretical analysis is presented, demonstrating how to derive optimal parameters for automating the method. It is then validated on both simulated data and very high resolution remote sensing imagery. The results show a clear improvement in change detection using the proposed method compared to other state-of-the-art change detection techniques.

We present an unsupervised hierarchical segmentation algorithm for detection of complex heterogeneous image structures that are comprised of simpler homogeneous primitive objects. An initial segmentation step produces regions corresponding to primitive objects with uniform spectral content. Next, the transitions between neighboring regions are modeled and clustered. We assume that the clusters that are dense and large enough in this transition space can be considered as significant. Then, the neighboring regions belonging to the significant clusters are merged to obtain the next level in the hierarchy. The experiments show that the algorithm that iteratively clusters and merges region groups is able to segment high-level complex structures in a hierarchical manner.

This paper extends a recent image-dependent regularisation approach introduced in aiming at edge-preserving smoothing. For that purpose, geodesic distances equipped with a Riemannian metric need to be estimated in local neighbourhoods. By deriving an appropriate metric from the gradient structure tensor, the associated geodesic paths are constrained to follow salient features in images. Following, we design a generalised anisotropic geodesic filter, incorporating not only a measure of the edge strength, like in the original method, but also further directional information about the image structures. The proposed filter is particularly efficient at smoothing heterogeneous areas while preserving relevant structures in multichannel images.

In this paper, we present a simple and heuristic ramp sharpening algorithm that achieves local contrast enhancement of vector-valued images. The proposed algorithm performs pixel wise comparisons of intensity values, gradient strength and directional information in order to locate transition ramps around true edges in the image. The sharpening is then applied only for those pixels found on the ramps. This way, the contrast between objects and regions separated by a ramp is enhanced correspondingly, avoiding ringing artifacts. It is found that applying this technique in an iterative manner on blurred imagery produces sharpening preserving both structure and signature of the image. The final approach reaches a good compromise between complexity and effectiveness for image simplfication, enhancing in an efficient manner the image details and maintaining the overall image appearance.

In order to describe, to extract image information content, segmentation is a well-known approach to represent the information in terms of objects. Image segmentation is a common image processing technique aiming at disintegrating an image into a partition of its support. Hierarchical of fuzzy segmentation are extension of segmentation definition, in order to provide a covering of the image support with overlapping segments. In this paper, we propose a novel approach for breaking up an image into multi-scale overlapping objects. The image is decomposed by granulometry or differential morphological pyramid, resulting in a discrete scale-space representation. Then, the scale-space transform is segmented by a region based method. Projecting the obtained scale-space partition into space constitutes the disintegrated image representation, which enables a multi-scale object based image description.

In this paper a new feature descriptor, the differential area profile (DAP), is presented. DAPs, like the regular differential morphological profiles, are computed from some size distribution. The proposed method is based on the area metric given by regular connected area filters. Area compared to local width, i.e. the diameter of the structuring element in the corresponding set of openings by reconstruction in classical DMPs, leads to a rather different multi-scale decomposition. This is investigated here and an example on a very high resolution satellite image tile is given.

Recent experience in using meter and sub-meter satellite images in support to crisis management is discussed in this paper. Technical challenges and conceptual gaps in automatic information extraction are described from a perspective of the Joint Research Center's experience in crisis management. Some methodological propositions are given to cope with these challenges, which were revealed in experiments realized in various crisis scenarios.

This paper introduces an image decomposition and simplification method based on the constrained connectivity paradigm. According to this paradigm, two pixels are said to be connected if they comply to a series of constraints defined in terms of simple measures such as the maximum grey level differences over well-defined pixel paths and regions. The resulting connectivity relation generates a unique partition of the image definition domain. The simplification of the image is then achieved by setting each segment of the partition to the mean value of the pixels falling within this segment. Fine to coarse partition hierarchies (and therefore images of increasing degree of simplification) are produced by varying the threshold value associated with each connectivity constraint. The paper also includes a generalisation to multichannel images, applications, a review of related image segmentation techniques, and pseudo-code for an implementation based on queue and stack data structures.

Users of remotely sensed imagery are often faced with the need to stitch two or more overlapping scenes together. This is generally referred to as image composition and the outcome is a single image with no overlapping regions which is also called an image mosaic. The most difficult part of image composition however, is deciding where to place the cut line in overlapping regions. In the days of hardcopy remote sensing, technicians would cut pictures along salient image structures or features when creating mosaics in order to minimise the ability of visually detecting the cut lines. Recently, a morphological image compositing algorithm was proposed that is able to automatically delineate cut lines along salient image structures. This technique was developed specifically to deal with very large image data sets automatically and generates a visually appealing image mosaic. To minimise the visual detection of a cut line, the contrast should look natural and therefore cuts should not be made through image objects but rather along the object. This paper presents the algorithm behind the morphological image compositing technique and how it was applied to automatically generate a European wide image mosaic based on over 800 Landsat ETM+ scenes. A quantitative measure was developed in order to try and quantify the 'quality' of an automatically delineated seam line. The results demonstrate how invisible the delineated seam lines are to the human eye based on the morphological image compositing method but at the same time, the experiments show the difficulty of quantitatively determining the 'best' one.

The goal of this paper is to segment high-resolution images of natural landscapes into different cover types. With this aim, morphological texture features (descriptors of random sets obtained by morphological transformations) are used in order to avoid the limitations of spectral features. First, a supervised segmentation (the textures to detect having been previously determined) is presented. The classes correspond to different degrees of tree densities. Second, a methodology for an unsupervised texture segmentation (no a priori information about the textures is supplied) is proposed. The number of classes is automatically determined. The proposed procedures have been tested on several images, providing promising results

Image mosaicking can be defined as the registration of two or more images that are then combined into a single image. Once the images have been registered to a common coordinate system, the problem amounts to the definition of a selection rule to output a unique value for all those pixels that are present in more than one image. This process is known as image compositing. In this paper, we propose a compositing procedure based on mathematical morphology and its marker-controlled segmentation paradigm. Its scope is to position seams along salient image structures so as to diminish their visibility in the output mosaic even in the absence of radiometric corrections or blending procedures. We also show that it is suited to the seamless minimization of undesirable transient objects occurring in the regions where two or more images overlap. The proposed methodology and algorithms are illustrated for the composition of satellite images minimizing cloud cover.

The goal of this work is to segment high resolution images of natural landscapes into different cover types. With this aim, morphological texture descriptors are used in order to avoid the limitations of spectral features. Firstly, an unsupervised texture segmentation method is presented, since no a priori information about the textures is supplied. Secondly, the same features are utilized in a supervised case, where the textures to detect have been previously determined.

By concentrating on the analysis of the spatial relationships between groups of pixels, mathematical morphology provides us with an image processing strategy complementary to those based on the analysis of the spectral signature of single pixels. A wide variety of morphological transformations are available for extracting structural information in spatial data. Accordingly, a stream of successful applications in geoscience and remote sensing have been reported since the mid-1980s as highlighted in a brief survey. However, recent advances in the theory of mathematical morphology still remain largely unexplored. We show in this paper that they can enhance methodologies for the processing and analysis of Earth observation data for tasks as diverse as filtering, simplification, directional segmentation and crest line extraction. We also address important issues overlooked in the past and concerning the applicability of a given morphological filter to Earth observation data. In particular, we point out that self-dual or even self-complementary filters are required in many applications to produce results independent of the local contrast of the searched image structures.

We show that a translation invariant implementation of min/max filters along a line segment of slope in the form of an irreducible fraction dy/dx can be achieved at the cost of 2+k min/max comparisons per image pixel, where k=max(|dx|,|dy|). Therefore, for a given slope, the computation time is constant and independent of the length of the line segment. We then present the notion of periodic moving histogram algorithm. This allows for a similar performance to be achieved in the more general case of rank filters and rank-based morphological filters. Applications to the filtering of thin nets and computation of both granulometries and orientation fields are detailed. Finally, two extensions are developed. The first deals with the decomposition of discrete disks and arbitrarily oriented discrete rectangles, while the second concerns min/max filters along gray tone periodic line segments.

The convex hull of a set is the smallest convex set containing this set. In ℜ2 , this definition is equivalent to the intersection of all half-planes containing the set. We show that this latter definition is nothing but an algebraic closing that can

New morphological tools for investigating local and global image structure orientations in grey-scale images are proposed. These tools are based on recent advances in the computation of erosions and dilations with line segments in arbitrary directions.

Van Herk (1992) has shown that the erosion/dilation operator with a linear structuring element of an arbitrary length can be implemented in only three min/max operations per pixel. In this paper, the algorithm is generalized to erosions and dilations along discrete lines at arbitrary angles. We also address the padding problem; so that the operation can be performed in place without copying the pixels to and from an intermediate buffer. Applications to image filtering and to radial decompositions of discs are presented.

To validate models of sheet metal forming processes, experiments with sheets of metal marked with grids are performed. A method for measuring the deformed grid after processing is presented. It is based on morphological image processing and it requires four steps. A large top hat transformation is first applied to homogenize the grey level intensities of the background. The two main directions of the grid are then determined by calculating covariances. Linear filtering along the directions of the grid is used to extract a mask of the grid pattern. This mask is then skeletonized in order to extract the medial axis of the grid as well as its intersection points.>

A fast and flexible algorithm for computing watersheds in digital gray-scale images is introduced. A review of watersheds and related motion is first presented, and the major methods to determine watersheds are discussed. The algorithm is based on an immersion process analogy, in which the flooding of the water in the picture is efficiently simulated using of queue of pixel. It is described in detail provided in a pseudo C language. The accuracy of this algorithm is proven to be superior to that of the existing implementations, and it is shown that its adaptation to any kind of digital grid and its generalization to n-dimensional images (and even to graphs) are straightforward. The algorithm is reported to be faster than any other watershed algorithm. Applications of this algorithm with regard to picture segmentation are presented for magnetic resonance (MR) imagery and for digital elevation models. An example of 3-D watershed is also provided.>

The watershed transformation is a very powerful image analysis tool provided by mathematical morphology. However, most existing watershed algorithms are either too time consuming or insufficiently accurate. The purpose of this paper is to introduce a new and flexible implementation of this transformation. It is based on a progressive flooding of the picture and it works for n-dimensional images. Pixels are first sorted in the increasing order of their gray values. Then, the successive gray levels are processed in order to simulate the flooding propagation. A distributive sorting technique combined with breadth-first scannings of each gray level allow an extremely fast computation. Furthermore, the present algorithm is very general since it deals with any kind of digital grid and its extension to general graphs is straightforward. Its interest with respect to image segmentation is illustrated by the extraction of geometrical shapes from a noisy image, the separation of 3-dimensional overlapping particles and by the segmentation of a digital elevation model using watersheds on images and graphs.