The Experts below are selected from a list of 35691 Experts worldwide ranked by ideXlab platform
Jianhua Gong - One of the best experts on this subject based on the ideXlab platform.
-
UAV Remote sensing for urban Vegetation Mapping using random forest and texture analysis
Remote Sensing, 2015Co-Authors: Quanlong Feng, Jiantao Liu, Jianhua GongAbstract:Unmanned aerial vehicle (UAV) remote sensing has great potential for Vegetation Mapping in complex urban landscapes due to the ultra-high resolution imagery acquired at low altitudes. Because of payload capacity restrictions, off-the-shelf digital cameras are widely used on medium and small sized UAVs. The limitation of low spectral resolution in digital cameras for Vegetation Mapping can be reduced by incorporating texture features and robust classifiers. Random Forest has been widely used in satellite remote sensing applications, but its usage in UAV image classification has not been well documented. The objectives of this paper were to propose a hybrid method using Random Forest and texture analysis to accurately differentiate land covers of urban vegetated areas, and analyze how classification accuracy changes with texture window size. Six least correlated second-order texture measures were calculated at nine different window sizes and added to original Red-Green-Blue (RGB) images as ancillary data. A Random Forest classifier consisting of 200 decision trees was used for classification in the spectral-textural feature space. Results indicated the following: (1) Random Forest outperformed traditional Maximum Likelihood classifier and showed similar performance to object-based image analysis in urban Vegetation classification; (2) the inclusion of texture features improved classification accuracy significantly; (3) classification accuracy followed an inverted U relationship with texture window size. The results demonstrate that UAV provides an efficient and ideal platform for urban Vegetation Mapping. The hybrid method proposed in this paper shows good performance in differentiating urban Vegetation Mapping. The drawbacks of off-the-shelf digital cameras can be reduced by adopting Random Forest and texture analysis at the same time.
Caiyun Zhang - One of the best experts on this subject based on the ideXlab platform.
-
A Framework to Combine Three Remotely Sensed Data Sources for Vegetation Mapping in the Central Florida Everglades
Wetlands, 2016Co-Authors: Caiyun Zhang, Donna Selch, Hannah CooperAbstract:A framework was designed to integrate three complimentary remotely sensed data sources (aerial photography, hyperspectral imagery, and LiDAR) for Mapping Vegetation in the Florida Everglades. An object-based pixel/feature-level fusion scheme was developed to combine the three data sources, and a decision-level fusion strategy was applied to produce the final Vegetation map by ensemble analysis of three classifiers k -Nearest Neighbor ( k -NN), Support Vector Machine (SVM), and Random Forest (RF). The framework was tested to map 11 land-use/land-cover level Vegetation types in a portion of the central Florida Everglades. An informative and accurate Vegetation map was produced with an overall accuracy of 91.1 % and Kappa value of 0.89. A combination of the three data sources achieved the best result compared with applying aerial photography alone, or a synergy of two data sources. Ensemble analysis of three classifiers not only increased the classification accuracy, but also generated a complementary uncertainty map for the final classified Vegetation map. This uncertainty map was able to identify regions with a high robust classification, as well as areas where classification errors were most likely to occur.
-
combining hyperspectral and lidar data for Vegetation Mapping in the florida everglades
Photogrammetric Engineering and Remote Sensing, 2014Co-Authors: Caiyun ZhangAbstract:This study explored a combination of hyperspectral and lidar systems for Vegetation Mapping in the Florida Everglades. A framework was designed to integrate two remotely sensed datasets and four data processing techniques. Lidar elevation and intensity features were extracted from the original point cloud data to avoid the errors and uncertainties in the raster-based lidar methods. Lidar significantly increased the classification accuracy compared with the application of hyperspectral data alone. Three lidar-derived features (elevation, intensity, and topography) had the same contributions in the classification. A synergy of hyperspectral imagery with all lidar-derived features achieved the best result with an overall accuracy of 86 percent and a Kappa value of 0.82 based on an ensemble analysis of three machine learning classifiers. Ensemble analysis did not signifi cantly increase the classification accuracy, but it provided a complementary uncertainty map for the final classified map. The study shows the promise of the synergy of hyperspectral and lidar systems for Mapping complex wetlands.
-
data fusion and classifier ensemble techniques for Vegetation Mapping in the coastal everglades
Geocarto International, 2014Co-Authors: Caiyun Zhang, Zhixiao XieAbstract:This study examined the applicability of data fusion and classifier ensemble techniques for Vegetation Mapping in the coastal Everglades. A framework was designed to combine these two techniques. In the framework, 20-m hyperspectral imagery collected from Airborne Visible/Infrared Imaging Spectrometer was first merged with 1-m Digital Orthophoto Quarter Quads using a proposed pixel/feature-level fusion strategy. The fused data set was then classified with an ensemble approach based on two contemporary machine learning algorithms: Random Forest and Support Vector Machine. The framework was applied to classify nine Vegetation types in a portion of the coastal Everglades. An object-based Vegetation map was produced with an overall accuracy of 90% and Kappa value of 0.86. Per-class classification accuracy varied from 61% for identifying buttonwood forest to 100% for identifying red mangrove scrub. The result shows that the framework is promising for automated Vegetation Mapping in the Everglades.
-
Object-based Vegetation Mapping in the Kissimmee River Watershed Using HyMap Data and Machine Learning Techniques
Wetlands, 2013Co-Authors: Caiyun ZhangAbstract:Accurate and informative Vegetation maps are in urgent demand to support the Kissimmee-Okeechobee-Everglades ecosystem restoration project in South Florida. In this study, we evaluated the applicability of fine spatial resolution hyperspectral data collected from the HyMap sensor for both community- and species-level Vegetation Mapping. Informative and accurate Vegetation maps were produced by combining machine learning methods (Support Vector Machines (SVM) and Random Forest (RF)), object-based image analysis techniques, and Minimum Noise Fraction (MNF) data transformation. An overall accuracy of 90% was obtained in discriminating 14 Vegetation communities. Classification of a large number of species is also promising. An overall accuracy of 85% was achieved in identifying 55 species using a SVM model. The results indicate that fine spatial resolution hyperspectral data classification using such automated procedure has great potential to replace the manual interpretation of aerial photos for Vegetation Mapping in heterogeneous wetland ecosystems.
-
combining object based texture measures with a neural network for Vegetation Mapping in the everglades from hyperspectral imagery
Remote Sensing of Environment, 2012Co-Authors: Caiyun ZhangAbstract:article i nfo An informative and accurate Vegetation map for the Greater Everglades of South Florida is in an urgent need to assist with the Comprehensive Everglades Restoration Plan (CERP), a $10.5-billion mission to restore the south Florida ecosystem in 30+ years. In this study, we examined the capability of fine spatial resolution hyperspectral imagery collected from Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) for vegeta- tion Mapping in the Everglades. In order to obtain an efficient and accurate procedure for Vegetation discrim- ination, we developed a neural network classifier first and then combined the object-based texture measures with the classifier to examine the contribution of the spatial information for Vegetation Mapping. The neural network is capable of modeling the characteristics of multiple spectral and spatial signatures within a class by an internally unsupervised engine and characterizing spectral and spatial differences between classes by an externally supervised system. The designed procedure was tested in a portion of the Everglades. An object- based Vegetation map was generated with an overall classification accuracy averaged 94% and a kappa value averaged 0.94 in discriminating 15 classes. The results are significantly better than those obtained from conventional classifiers such as maximum likelihood and spectral angle mapper. The study illustrates that combining object-based texture measures in the neural network classifier can significantly improve the classification. It is concluded that fine spatial resolution hyperspectral data is an effective solution to ac- curate Vegetation Mapping in the Everglades which has a rich plant community with a high degree of spatial and spectral heterogeneity.
Quanlong Feng - One of the best experts on this subject based on the ideXlab platform.
-
UAV Remote sensing for urban Vegetation Mapping using random forest and texture analysis
Remote Sensing, 2015Co-Authors: Quanlong Feng, Jiantao Liu, Jianhua GongAbstract:Unmanned aerial vehicle (UAV) remote sensing has great potential for Vegetation Mapping in complex urban landscapes due to the ultra-high resolution imagery acquired at low altitudes. Because of payload capacity restrictions, off-the-shelf digital cameras are widely used on medium and small sized UAVs. The limitation of low spectral resolution in digital cameras for Vegetation Mapping can be reduced by incorporating texture features and robust classifiers. Random Forest has been widely used in satellite remote sensing applications, but its usage in UAV image classification has not been well documented. The objectives of this paper were to propose a hybrid method using Random Forest and texture analysis to accurately differentiate land covers of urban vegetated areas, and analyze how classification accuracy changes with texture window size. Six least correlated second-order texture measures were calculated at nine different window sizes and added to original Red-Green-Blue (RGB) images as ancillary data. A Random Forest classifier consisting of 200 decision trees was used for classification in the spectral-textural feature space. Results indicated the following: (1) Random Forest outperformed traditional Maximum Likelihood classifier and showed similar performance to object-based image analysis in urban Vegetation classification; (2) the inclusion of texture features improved classification accuracy significantly; (3) classification accuracy followed an inverted U relationship with texture window size. The results demonstrate that UAV provides an efficient and ideal platform for urban Vegetation Mapping. The hybrid method proposed in this paper shows good performance in differentiating urban Vegetation Mapping. The drawbacks of off-the-shelf digital cameras can be reduced by adopting Random Forest and texture analysis at the same time.
Laurence Hubert-moy - One of the best experts on this subject based on the ideXlab platform.
-
Evaluation of bispectral LIDAR data for urban Vegetation Mapping
Proceedings SPIE, 2016Co-Authors: Jean Nabucet, Thomas Corpetti, Laurence Hubert-moy, Patrick Launeau, Dimitri Lague, Cyril Michon, Hervé QuénolAbstract:Because of the large increase of urban population in the last decades, the question of sustainable development in urban areas is crucial. In this context, Vegetation plays a significant role in urban planning, environmental protecting, and sustainable development policy making, heating and cooling requirements of buildings, displacement of animals dispersion, concentration of pollutants, and well-being. In numerous cities, Vegetation is limited to public areas using GPS surveys or aerial remote sensing data. Recently, very high-resolution sensors as Light Detection and Ranging (LiDAR) data have permitted significant improvements in Vegetation Mapping in urban areas. This paper presents an evaluation of a new generation of airborne LIDAR bi-spectral discrete point (Optech titan) for Mapping and characterizing urban Vegetation. The methodology is based on a four-step approach: 1) the analysis of the quality of data in order to estimate noise between the green and near-infrared LIDAR point clouds; 2) this enables to remove the topographic effects and 3) a first classification, devoted to the elimination of the non-Vegetation class, is performed based on the intensity value of the two channels; finally, in 4), the tree coverage is classified into seven categories of strata combination. To this end specific descriptors related to the organization of the point clouds are used. These first results show that compared to monospectral LiDAR data, bi-spectral LiDAR enables to improve significantly both the extraction and the characterization of urban objects. This reveals new perspectives for Mapping and characterizing urban patterns and other complex structures. © (2016) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
-
Multiscale comparison of remote-sensing data for linear woody Vegetation Mapping
International Journal of Remote Sensing, 2014Co-Authors: C. Vannier, Laurence Hubert-moyAbstract:Wooded hedgerows do not cover large areas but perform many functions that are beneficial to water quality and biodiversity. A broad range of remotely sensed data is available to map these small linear elements in rural landscapes, but only a few of them have been evaluated for this purpose. In this study, we evaluate and compare various optical remote-sensing data including high and very high spatial resolution, active and passive, and airborne and satellite data to produce quantitative information on the hedgerow network structure and to analyse qualitative information from the maps produced in order to estimate the true value of these maps. We used an object-based image analysis that proved to be efficient for detecting and Mapping thin elements in complex landscapes. The analysis was performed at two scales, the hedgerow network scale and the tree canopy scale, on a study site that shows a strong landscape gradient of wooded hedgerow density. The results (1) highlight the key role of spectral resolution on the detection and Mapping of wooded elements with remotely sensed data; (2) underline the fact that every satellite image provides relevant information on wooded network structures, even in closed landscape units, whatever the spatial resolution; and (3) indicate that light detection and ranging data offer important insights into future strategies for monitoring hedgerows.
-
Multi-temporal classification of TerraSAR-X data for wetland Vegetation Mapping
2013Co-Authors: Julie Betbeder, Eric Pottier, Sebastien Rapinel, Thomas Corpetti, Samuel Corgne, Laurence Hubert-moyAbstract:This paper is concerned with Vegetation wetland Mapping using multi-temporal SAR imagery. Whilst wetlands play a key role in controlling flooding and nonpoint source pollution, sequestering carbon and providing an abundance of ecological services, knowledge of the flora and fauna of these environments is patchy, and understanding of their ecological functioning is still insufficient for a reliable functional assessment on areas larger than a few ha. The aim of this paper is to evaluate multitemporal TerraSAR-X imagery to map precisely the distribution of Vegetation formations within wetlands, in determining seasonally flooded areas of wetlands. A series of six dual-polarization TerraSAR-X images (HH/VV) were acquired in 2012 during dry and wet seasons. Polarimetric and intensity parameters, which present a temporal variation that depends on wetland flooding status and Vegetation roughness, were firstly extracted. The parameters were then classified based on Support Vector Machines (SVM) techniques using a specific kernel adapted to the comparison of time-series data. The results show that the Shannon entropy parameter allows discriminating Vegetation formations within wetland with more accuracy than intensity parameters.
Damir Medak - One of the best experts on this subject based on the ideXlab platform.
-
Spatial video remote sensing for urban Vegetation Mapping using Vegetation indices
Urban Ecosystems, 2020Co-Authors: Luka Rumora, Ivan Majić, Mario Miler, Damir MedakAbstract:Urban Vegetation is important because of the fast-growing urbanization. If we want cities to have sustainable growth and well-kept ecology, we need to develop a smart and efficient urban Vegetation monitoring system. This paper examines the possibility of using a modified GoPro camera mounted on a car. The lens of a GoPro camera was replaced with the NDVI-7 lens to obtain blue, green and near-infrared band. The performance of four Vegetation indices was tested: Blue normalized difference Vegetation index (BNDVI), Green normalized difference Vegetation index (GNDVI), Green-blue normalized difference Vegetation index (GBNDVI), Blue-wide dynamic range Vegetation index (BWDRVI). Based on those indices, binary classification was performed to classify objects in the scene as either Vegetation or non-Vegetation. Finally, the accuracy of each index was assessed on three different study sites. Results show that GBNDVI performs best for the given task with average classification accuracy of 95.10% for all study sites.