199 research outputs found
Agricultural Field Boundary Data Set in The Netherlands for Automatic Delineation using Sentinel-2 Images
No full textThe data set contains Sentinel-2 image tiles and corresponding reference maps derived from PDOK for agricultural field boundaries. Predictions using three different deep learning models are included
Sentinel-2 data set for the delineation of agricultural field boundaries in Flevoland, The Netherlands
No full textSentinel-2 data set for the delineation of agricultural boundaries. It contains filed boundaries in raster format at 10 m and 5 m resolution and in vector format. It contains Sentinel-2 bands at 10m and 20 m resolution. RapydEye images at 5 m resolution are also available, for validation puposes
A Novel Context-Sensitive SVM for Classification of Remote Sensing Images
No full textIn this paper, a novel context-sensitive classification technique based on Support Vector Machines (CS-SVM) is proposed. This technique aims at exploiting the promising SVM method for classification of 2-D (or n-D) scenes by considering the spatial-context information of the pixel to be analyzed. In greater detail, the proposed architecture properly exploits the spatial-context information for: i) increasing the robustness of the learning procedure of SVMs to the noise present in the training set (mislabeled training samples); ii) regularizing the classification maps. The first property is achieved by introducing a context-sensitive term in the objective function to be minimized for defining the decision hyperplane in the SVM kernel space. The second property is obtained including in the classification procedure of a generic pattern the information of neighboring pixels. Experiments carried out on very high geometrical resolution images confirm the validity of the proposed technique
Cost-sensitive active learning with lookahead: optimizing field surveys for remote sensing data classification
No full textActive learning typically aims at minimizing the number of labeled samples to be included in the training set to reach a certain level of classification accuracy. Standard methods do not usually take into account the real annotation procedures and implicitly assume that all samples require the same effort to be labeled. Here, we consider the case where the cost associated with the annotation of a given sample depends on the previously labeled samples. In general, this is the case when annotating a queried sample is an action that changes the state of a dynamic system, and the cost is a function of the state of the system. In order to minimize the total annotation cost, the active sample selection problem is addressed in the framework of a Markov decision process, which allows one to plan the next labeling action on the basis of an expected long-term cumulative reward. This framework allows us to address the problem of optimizing the collection of labeled samples by field surveys for the classification of remote sensing data. The proposed method is applied to the ground sample collection for tree species classification using airborne hyperspectral images. Experiments carried out in the context of a real case study on forest inventory show the effectiveness of the proposed metho
Optimizing the ground sample collection with cost-sensitive active learning for tree species classification using hyperspectral images
No full textThis study presents a cost-sensitive active learning method for optimizing the field surveys by a human expert in the classification of single tree species using hyperspectral images. The goal of the proposed method is to guide the human expert in the collection of labeled samples in order to maximize the ratio between the classification accuracy with respect to the
travelling costs. Experiments carried out in the context of a real study on forest inventory show the effectiveness of the
proposed metho
From colloidal dispersions to colloidal pastesthrough solid–liquid separation processes
No full textSolid–liquid separation is an operation that starts with a dispersion of solid particles in a liquid and removes some of the liquid from the particles, producing a concentrated solid paste and a clean liquid phase. It is similar to thermodynamic processes where pressure is applied to a system in order to reduce its volume. In dispersions, the resistance to this osmotic compression depends on interactions between the dispersed particles. The first part of this work deals with dispersions of repelling particles, which are either silica nanoparticles or synthetic clay platelets, dispersed in aqueous solutions. In these conditions, each particle is surrounded by an ionic layer, which repels other ionic layers. This results in a structure with strong short-range order. At high particle volume fractions, the overlap of ionic layers generates large osmotic pressures; these pressures may be calculated, through the cell model, as the cost of reducing the volume of each cell. The variation of osmotic pressure with volume fraction is the equation of state of the dispersion. The second part of this work deals with dispersions of aggregated particles, which are silica nanoparticles, dispersed in water and flocculated by multivalent cations. This produces large bushy aggregates, with fractal structures that are maintained through interparticle surface– surface bonds. As the paste is submitted to osmotic pressures, small relative displacements of the aggregated particles lead to structural collapse. The final structure is made of a dense skeleton immersed in a nearly homogeneous matrix of aggregated particles. The variation of osmotic resistance with volume fraction is the compression law of the paste; it may be calculated through a numerical model that takes into account the noncentral interparticle forces. According to this model, the response of aggregated pastes to applied stress may be controlled through the manipulation of interparticle adhesion
Agricultural Field Boundary Data Set in The Netherlands for Automatic Delineation using Sentinel-2 Images
No full textThe data set contains Sentinel-2 image tiles and corresponding reference maps derived from PDOK for agricultural field boundaries. Predictions using three different deep learning models are included
Sentinel-2 data set for the delineation of agricultural field boundaries in Flevoland, The Netherlands
No full textSentinel-2 data set for the delineation of agricultural boundaries. It contains filed boundaries in raster format at 10 m and 5 m resolution and in vector format. It contains Sentinel-2 bands at 10m and 20 m resolution. RapydEye images at 5 m resolution are also available, for validation puposes
Change detection image dataset for unplanned settlement in Rwanda
No full textUnmanned Aerial Vehicle imagery from 2015 and 2017 over an unplanned settlement in Kigali, Rwanda. These images were used for the change detection studies published in the article:Gevaert, C. M., Persello, C., Sliuzas, R. V., & Vosselman, G. (2020). Monitoring household upgrading in unplanned settlements with unmanned aerial vehicles. International Journal of Applied Earth Observation and Geoinformation (JAG), 90, 1-6. [102117]. https://doi.org/10.1016/j.jag.2020.102117</p
- …
