This paper proposes a new framework for ground extraction and building detection in LiDAR data. The proposed approach constructs the connectivity of a grid over the LiDAR point-cloud in order to perform multi-scale data decomposition. This is realised by forming a top-hat scale-space using differential morphological profiles (DMPs) on pointsʼ residuals from the approximated surface. The geometric attributes of the contained features are estimated by mapping characteristic values from DMPs. Ground definition is achieved by using featuresʼ geometry, whilst their surface and regional attributes are additionally considered for building detection. A new algorithmfor local fitting surfaces (LoFS) is proposed for extracting planar points. Finally, transitions between planar ground and non-ground regions are observed in order to separate regions of similar geometrical and surface properties but different contexts (i.e. bridges and buildings). The methods were evaluated using ISPRS benchmark datasets and show superior results in comparison to the current state-of-the-art.
COBISS.SI-ID: 17466134
This paper proposes a new method for the detection of vegetation in LiDAR data. As vegetation points are characterised by non-linear distributions, they are efficiently recognised based-on large errors obtained when applying the local fitting of planar surfaces. In addition, three contextual filters are introduced capable of dealing with those exceptions that do not conform with previous interpretations. Namely, they are designed for detecting overgrowing vegetation, small objects attached to the planar surfaces (such as balconies, chimneys, and noise within the buildings) and small objects that do not belong to vegetation (vehicles, statues, fences). During the validation, the proposed method achieved over 97% correctness as well as completeness of vegetation recognition in rural areas while its average accuracy in urban settings was 90.7% in terms of F1F1-scores. The method uses only three input parameters and allows for efficient compensation between completeness and correctness, without significantly affecting the F1F1-score. Sensitivity analysis of the method also confirmed the robustness against a sub-optimal definition of the input parameters.
COBISS.SI-ID: 19409174
In this paper a time series of small glacier volume measurements performed with ground penetrating radar measurements (GPR) is presented. Volume changes of small glaciers in mid latitudes and especially those located at low altitude are important because they respond suddenly to climate changes both on local and global scale. We present a GPR dataset acquired on September 23 and 24, 2013 on the Triglav glacier to identify layers with different characteristics (snow, firn, ice, debris) within the glacier and to define the extension and volume of the actual ice. Here we compare also the results with a previous GPR survey acquired in 2000. A critical review of the historical data to find the general trend and to forecast a possible evolution is also presented. Between 2000 and 2013, we observed relevant changes in the internal distribution of the different units (snow, firn, ice) and the ice volume reduced from about 35,000 m3 to about 7400 m3.
COBISS.SI-ID: 40064557
An important problem in the analysis of network data is the detection of groups of densely interconnected nodes also called modules or communities. Community structure reveals functions and organizations of networks. Currently used algorithms for community detection in large-scale realworld networks are computationally expensive or require a priori information such as the number or sizes of communities or are not able to give the same resulting partition in multiple runs. In this paper we investigate a simple and fast algorithm that uses the network structure alone and requires neither optimization of pre-defined objective function nor information about number of communities. We propose a bottom up community detection algorithm in which starting from communities consisting of adjacent pairs of nodes and their maximal similar neighbors we find real communities. We show that the overall advantage of the proposed algorithm compared to the other community detection algorithms is its simple nature, low computational cost and its very high accuracy in detection communities of different sizes also in networks with blurred modularity structure consisting of poorly separated communities. All communities identified by the proposed method for facebook network and E-Coli transcriptional regulatory network have strong structural and functional coherence.
COBISS.SI-ID: 21987592
This paper proposes a new mapping schema, named mapping, for filtering nonground objects fromLiDAR data, and the generation of a digital terrain model. By extending the CSL model, mapping extracts the most contrasted connected-components from top-hat scale-space and attributes them for an adaptive multicriterion filter definition. Areas of the most contrasted connected-components and the standard deviations of contained pointsʼ levels are considered for this purpose. Computational efficiency is achieved by arranging the input LiDAR data into a grid, represented by a Max-Tree. Since aconstant number of passes over the grid is required, the time complexity of the proposed method is linear according to the number of grid-cells. As confirmed by the experiments, the average CPU execution time decreases by nearly 98%, while the average accuracy improves by up to 10% in comparison with the related method.
COBISS.SI-ID: 16937494