Geo-Spatial Support for Assessment of Anthropic Impact  on Biodiversity

Piragnolo, Marco; Pirotti, Francesco; Guarnieri, Alberto; Vettore, Antonio; Salogni, G.

doi:10.3390/ijgi3020599

Cited by 22 publications

(21 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the urban context, fitting models can help to contribute to the "smart-city" paradigm, e.g. by monitoring land-surface temperature (Scarano, 2015) or providing data for anthropic impact assessment in urban areas and outside urban areas (Akın et.al., 2015;Piragnolo et al, 2014). In environmental context, remote sensing provides a global view of the Earth's phenomena and all the variables which are necessary to assess and predict its dynamics.…”

Section: Introductionmentioning

confidence: 99%

Benchmark of Machine Learning Methods for Classification of a Sentinel-2 Image

Pirotti¹,

Sunar²,

Piragnolo³

2016

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

Commission VII, WG VII/4KEY WORDS: Machine learning, Sentinel-2, Remote sensing, Neural nets, Agriculture, Land cover, Classification ABSTRACT:Thanks to mainly ESA and USGS, a large bulk of free images of the Earth is readily available nowadays. One of the main goals of remote sensing is to label images according to a set of semantic categories, i.e. image classification. This is a very challenging issue since land cover of a specific class may present a large spatial and spectral variability and objects may appear at different scales and orientations. In this study, we report the results of benchmarking 9 machine learning algorithms tested for accuracy and speed in training and classification of land-cover classes in a Sentinel-2 dataset. The following machine learning methods (MLM) have been tested: linear discriminant analysis, k-nearest neighbour, random forests, support vector machines, multi layered perceptron, multi layered perceptron ensemble, ctree, boosting, logarithmic regression. The validation is carried out using a control dataset which consists of an independent classification in 11 land-cover classes of an area about 60 km 2 , obtained by manual visual interpretation of high resolution images (20 cm ground sampling distance) by experts. In this study five out of the eleven classes are used since the others have too few samples (pixels) for testing and validating subsets. The classes used are the following: (i) urban (ii) sowable areas (iii) water (iv) tree plantations (v) grasslands. Validation is carried out using three different approaches: (i) using pixels from the training dataset (train), (ii) using pixels from the training dataset and applying cross-validation with the k-fold method (kfold) and (iii) using all pixels from the control dataset. Five accuracy indices are calculated for the comparison between the values predicted with each model and control values over three sets of data: the training dataset (train), the whole control dataset (full) and with k-fold cross-validation (kfold) with ten folds. Results from validation of predictions of the whole dataset (full) show the random forests method with the highest values; kappa index ranging from 0.55 to 0.42 respectively with the most and least number pixels for training. The two neural networks (multi layered perceptron and its ensemble) and the support vector machines -with default radial basis function kernel -methods follow closely with comparable performance.

show abstract

Section: Introductionmentioning

confidence: 99%

Benchmark of Machine Learning Methods for Classification of a Sentinel-2 Image

Pirotti¹,

Sunar²,

Piragnolo³

2016

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

show abstract

“…GIS-supported applications can deploy damage maps by using various sources to infer damaged ground cover (Piragnolo et al, 2015(Piragnolo et al, , 2014Pirotti, 2010a). In this case, we used aerial imagery to estimate damage from windthrow in ten areas, getting a very positive result.…”

Section: Discussionmentioning

confidence: 99%

Kernel Feature Cross-Correlation for Unsupervised Quantification of Damage From Windthrow in Forests

Pirotti

Travaglini

Giannetti

et al. 2016

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

Commission VII, WG VII/1 KEY WORDS: Feature matching, Kernel normalized cross-correlation, Forest, Support Vector Machines ABSTRACT:In this study estimation of tree damage from a windthrow event using feature detection on RGB high resolution imagery is assessed. An accurate quantitative assessment of the damage in terms of volume is important and can be done by ground sampling, which is notably expensive and time-consuming, or by manual interpretation and analyses of aerial images. This latter manual method also requires an expert operator investing time to manually detect damaged trees and apply relation functions between measures and volume which are also error-prone. In the proposed method RGB images with 0.2 m ground sample distance are analysed using an adaptive template matching method. Ten images corresponding to ten separate study areas are tested. A 13x13 pixels kernel with a simplified linear-feature representation of a cylinder is applied at different rotation angles (from 0° to 170° at 10° steps). The higher values of the normalized cross-correlation (NCC) of all angles are recorded for each pixel for each image. Several features are tested: percentiles (75, 80, 85, 90, 95, 99, max) and sum and number of pixels with NCC above 0.55. Three regression methods are tested, multiple regression (mr), support vector machines (svm) with linear kernel and random forests. The first two methods gave the best results. The ground-truth was acquired by ground sampling, and total volumes of damaged trees are estimated for each of the 10 areas. Damaged volumes in the ten areas range from ~1.8 x10 2 m 3 to ~1.2x10 4 m 3 . Regression results show that smv regression method over the sum gives an R-squared of 0.92, a mean of absolute errors (MAE) of 255 m 3 and a relative absolute error (RAE) of 34% using leave-one-out cross validation from the 10 observations. These initial results are encouraging and support further investigations on more finely tuned kernel template metrics to define an unsupervised image analysis process to automatically assess forest damage from windthrow.

show abstract

“…More complex interactions with existing datasets related to land management, e.g. biodiversity (Piragnolo et al, 2014), or even collaborative portals which integrate FW data with other remote sensing data can bring added value to end users. Having a lot of data does not imply having a lot of information, especially if users cannot extract important components from such data.…”

Section: Discussionmentioning

confidence: 99%

Processing lidar waveform data for 3D visual assessment of forest environments

Pirotti

Guarnieri

Masiero

et al. 2014

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

ABSTRACT:The objective of this report is to present and discuss a work-flow for extracting, from full-waveform (FW) lidar data, formats which are compatible with common information systems (GIS) and statistical software packages. Full-waveform, specifically for forestry, got attention from the scientific community because a more in-depth analysis can add valuable information for classification and modelling of related variables (e.g. biomass). In order to assess if this is feasible and if the results are useful, the end-user has to deal with raw datasets from lidar sensors. In this study case we propose and test a work-flow which is implemented through a selfdeveloped software integrating ad-hoc C++ libraries and a graphical user interface for an easier approach by end-users. This software allows the user to add raw FW data and produce several products which can successively be easily imported in GIS or statistical software. To achieve this we used some state-of-the-art methods which have been extensively reported in literature and we discuss results and future developments. Results show that this software package can effectively work as a tool for linking raw FW data with forest-related spatial processing by providing punctual information directly derived from the FW data or area-based aggregated information for a more generalized description of the earth surface.

show abstract

Geo-Spatial Support for Assessment of Anthropic Impact on Biodiversity

Cited by 22 publications

References 20 publications

Benchmark of Machine Learning Methods for Classification of a Sentinel-2 Image

Benchmark of Machine Learning Methods for Classification of a Sentinel-2 Image

Kernel Feature Cross-Correlation for Unsupervised Quantification of Damage From Windthrow in Forests

Processing lidar waveform data for 3D visual assessment of forest environments

Contact Info

Product

Resources

About