Sensors and OBIA synergy for operational monitoring of surface water

Science.gov (United States)

Masson, Eric; Thenard, Lucas

2010-05-01

This contribution will focus on combining Object Based Image Analysis (i.e. OBIA with e-Cognition 8) and recent sensors (i.e. Spot 5 XS, Pan and ALOS Prism, Avnir2, Palsar) to address the technical feasibility for an operational monitoring of surface water. Three cases of river meandering (India), flood mapping (Nepal) and dam's seasonal water level monitoring (Morocco) using recent sensors will present various application of surface water monitoring. The operational aspect will be demonstrated either by sensor properties (i.e. spatial resolution and bandwidth), data acquisition properties (i.e. multi sensor, return period and near real-time acquisition) but also with OBIA algorithms (i.e. fusion of multi sensors / multi resolution data and batch processes). In the first case of river meandering (India) we will address multi sensor and multi date satellite acquisition to monitor the river bed mobility within a floodplain using an ALOS dataset. It will demonstrate the possibility of an operational monitoring system that helps the geomorphologist in the analysis of fluvial dynamic and sediment budget for high energy rivers. In the second case of flood mapping (Nepal) we will address near real time Palsar data acquisition at high spatial resolution to monitor and to map a flood extension. This ALOS sensor takes benefit both from SAR and L band properties (i.e. atmospheric transparency, day/night acquisition, low sensibility to surface wind). It's a real achievement compared to optical imagery or even other high resolution SAR properties (i.e. acquisition swath, bandwidth and data price). These advantages meet the operational needs set by crisis management of hydrological disasters but also for the implementation of flood risk management plans. The last case of dam surface water monitoring (Morocco) will address an important issue of water resource management in countries affected by water scarcity. In such countries water users have to cope with over exploitation

Object-oriented classification of land use in urban areas applying very high resolution satellite data

International Nuclear Information System (INIS)

Bauer, T.B.

2001-08-01

The availability of the new very high resolution satellite imagery will offer a wide range of new applications in the field of remote sensing. Information about actual land use is an important task for the management and planning in urban areas. High resolution satellite data will be an alternative to aerial photographs for updating and maintaining cartographic and geographic databases at reduced costs. The aim of the research is to formalize the visual interpretation procedure in order to automate the whole process. The assumption underlying this approach is that the land use functions can be distinguished on the basis of the differences in spatial distribution and pattern of land cover forms. Therefore a two-stage classification procedure is applied. In a first stage a land cover map is produced. In a second stage the morphological properties and spatial patterns of the land cover objects are analyzed with the structural analyzing and mapping system leading to a characterization and description of distinct urban land use categories. This information is then used for building a rule system that is implemented in a new commercial software tool called eCognition. An object-oriented classifier applies the rules to the land cover objects resulting in the required land use map. The potential of this method is demonstrated in a case study using IKONOS data covering a part of the metropolitan area of Vienna. (author)

Land Use and Land Cover (LULC) Change Detection in Islamabad and its Comparison with Capital Development Authority (CDA) 2006 Master Plan

Science.gov (United States)

Hasaan, Zahra

2016-07-01

Remote sensing is very useful for the production of land use and land cover statistics which can be beneficial to determine the distribution of land uses. Using remote sensing techniques to develop land use classification mapping is a convenient and detailed way to improve the selection of areas designed to agricultural, urban and/or industrial areas of a region. In Islamabad city and surrounding the land use has been changing, every day new developments (urban, industrial, commercial and agricultural) are emerging leading to decrease in vegetation cover. The purpose of this work was to develop the land use of Islamabad and its surrounding area that is an important natural resource. For this work the eCognition Developer 64 computer software was used to develop a land use classification using SPOT 5 image of year 2012. For image processing object-based classification technique was used and important land use features i.e. Vegetation cover, barren land, impervious surface, built up area and water bodies were extracted on the basis of object variation and compared the results with the CDA Master Plan. The great increase was found in built-up area and impervious surface area. On the other hand vegetation cover and barren area followed a declining trend. Accuracy assessment of classification yielded 92% accuracies of the final land cover land use maps. In addition these improved land cover/land use maps which are produced by remote sensing technique of class definition, meet the growing need of legend standardization.

GRAIN-SIZE MEASUREMENTS OF FLUVIAL GRAVEL BARS USING OBJECT-BASED IMAGE ANALYSIS

Directory of Open Access Journals (Sweden)

Pedro Castro

2018-01-01

Full Text Available Traditional techniques for classifying the average grain size in gravel bars require manual measurements of each grain diameter. Aiming productivity, more efficient methods have been developed by applying remote sensing techniques and digital image processing. This research proposes an Object-Based Image Analysis methodology to classify gravel bars in fluvial channels. First, the study evaluates the performance of multiresolution segmentation algorithm (available at the software eCognition Developer in performing shape recognition. The linear regression model was applied to assess the correlation between the gravels’ reference delineation and the gravels recognized by the segmentation algorithm. Furthermore, the supervised classification was validated by comparing the results with field data using the t-statistic test and the kappa index. Afterwards, the grain size distribution in gravel bars along the upper Bananeiras River, Brazil was mapped. The multiresolution segmentation results did not prove to be consistent with all the samples. Nonetheless, the P01 sample showed an R2 =0.82 for the diameter estimation and R2=0.45 the recognition of the eliptical ft. The t-statistic showed no significant difference in the efficiencies of the grain size classifications by the field survey data and the Object-based supervised classification (t = 2.133 for a significance level of 0.05. However, the kappa index was 0.54. The analysis of the both segmentation and classification results did not prove to be replicable.

Woody encroachment over 70 years in South African savannahs: overgrazing, global change or extinction aftershock?

Science.gov (United States)

Stevens, Nicola; Erasmus, B F N; Archibald, S; Bond, W J

2016-09-19

Woody encroachment in 'open' biomes like grasslands and savannahs is occurring globally. Both local and global drivers, including elevated CO2, have been implicated in these increases. The relative importance of different processes is unresolved as there are few multi-site, multi-land-use evaluations of woody plant encroachment. We measured 70 years of woody cover changes over a 1020 km(2) area covering four land uses (commercial ranching, conservation with elephants, conservation without elephants and communal rangelands) across a rainfall gradient in South African savannahs. Different directions of woody cover change would be expected for each different land use, unless a global factor is causing the increases. Woody cover change was measured between 1940 and 2010 using the aerial photo record. Detection of woody cover from each aerial photograph was automated using eCognitions' Object-based image analysis (OBIA). Woody cover doubled in all land uses across the rainfall gradient, except in conservation areas with elephants in low-rainfall savannahs. Woody cover in 2010 in low-rainfall savannahs frequently exceeded the maximum woody cover threshold predicted for African savannahs. The results indicate that a global factor, of which elevated CO2 is the likely candidate, may be driving encroachment. Elephants in low-rainfall savannahs prevent encroachment and localized megafaunal extinction is a probable additional cause of encroachment.This article is part of the themed issue 'Tropical grassy biomes: linking ecology, human use and conservation'. © 2016 The Author(s).

Classification of high resolution remote sensing image based on geo-ontology and conditional random fields

Science.gov (United States)

Hong, Liang

2013-10-01

The availability of high spatial resolution remote sensing data provides new opportunities for urban land-cover classification. More geometric details can be observed in the high resolution remote sensing image, Also Ground objects in the high resolution remote sensing image have displayed rich texture, structure, shape and hierarchical semantic characters. More landscape elements are represented by a small group of pixels. Recently years, the an object-based remote sensing analysis methodology is widely accepted and applied in high resolution remote sensing image processing. The classification method based on Geo-ontology and conditional random fields is presented in this paper. The proposed method is made up of four blocks: (1) the hierarchical ground objects semantic framework is constructed based on geoontology; (2) segmentation by mean-shift algorithm, which image objects are generated. And the mean-shift method is to get boundary preserved and spectrally homogeneous over-segmentation regions ;(3) the relations between the hierarchical ground objects semantic and over-segmentation regions are defined based on conditional random fields framework ;(4) the hierarchical classification results are obtained based on geo-ontology and conditional random fields. Finally, high-resolution remote sensed image data -GeoEye, is used to testify the performance of the presented method. And the experimental results have shown the superiority of this method to the eCognition method both on the effectively and accuracy, which implies it is suitable for the classification of high resolution remote sensing image.

Detecting Slums from Quick Bird Data in Pune Using AN Object Oriented Approach

Science.gov (United States)

Shekhar, S.

2012-07-01

We have been witnessing a gradual and steady transformation from a pre dominantly rural society to an urban society in India and by 2030, it will have more people living in urban than rural areas. Slums formed an integral part of Indian urbanisation as most of the Indian cities lack in basic needs of an acceptable life. Many efforts are being taken to improve their conditions. To carry out slum renewal programs and monitor its implementation, slum settlements should be recorded to obtain an adequate spatial data base. This can be only achieved through the analysis of remote sensing data with very high spatial resolution. Regarding the occurrences of settlement areas in the remote sensing data pixel-based approach on a high resolution image is unable to represent the heterogeneity of complex urban environments. Hence there is a need for sophisticated method and data for slum analysis. An attempt has been made to detect and discriminate the slums of Pune city by describing typical characteristics of these settlements, by using eCognition software from quick bird data on the basis of object oriented approach. Based on multi resolution segmentation, initial objects were created and further depend on texture, geometry and contextual characteristics of the image objects, they were classified into slums and non-slums. The developed rule base allowed the description of knowledge about phenomena clearly and easily using fuzzy membership functions and the described knowledge stored in the classification rule base led to the best classification with more than 80% accuracy.

Object-Based Classification as an Alternative Approach to the Traditional Pixel-Based Classification to Identify Potential Habitat of the Grasshopper Sparrow

Science.gov (United States)

Jobin, Benoît; Labrecque, Sandra; Grenier, Marcelle; Falardeau, Gilles

2008-01-01

The traditional method of identifying wildlife habitat distribution over large regions consists of pixel-based classification of satellite images into a suite of habitat classes used to select suitable habitat patches. Object-based classification is a new method that can achieve the same objective based on the segmentation of spectral bands of the image creating homogeneous polygons with regard to spatial or spectral characteristics. The segmentation algorithm does not solely rely on the single pixel value, but also on shape, texture, and pixel spatial continuity. The object-based classification is a knowledge base process where an interpretation key is developed using ground control points and objects are assigned to specific classes according to threshold values of determined spectral and/or spatial attributes. We developed a model using the eCognition software to identify suitable habitats for the Grasshopper Sparrow, a rare and declining species found in southwestern Québec. The model was developed in a region with known breeding sites and applied on other images covering adjacent regions where potential breeding habitats may be present. We were successful in locating potential habitats in areas where dairy farming prevailed but failed in an adjacent region covered by a distinct Landsat scene and dominated by annual crops. We discuss the added value of this method, such as the possibility to use the contextual information associated to objects and the ability to eliminate unsuitable areas in the segmentation and land cover classification processes, as well as technical and logistical constraints. A series of recommendations on the use of this method and on conservation issues of Grasshopper Sparrow habitat is also provided.

TREE CANOPY COVER MAPPING USING LiDAR IN URBAN BARANGAYS OF CEBU CITY, CENTRAL PHILIPPINES

Directory of Open Access Journals (Sweden)

J. A. Ejares

2016-06-01

Full Text Available This paper investigates tree canopy cover mapping of urban barangays (smallest administrative division in the Philippines in Cebu City using LiDAR (Light Detection and Ranging. Object-Based Image Analysis (OBIA was used to extract tree canopy cover. Multi-resolution segmentation and a series of assign-class algorithm in eCognition software was also performed to extract different land features. Contextual features of tree canopies such as height, area, roundness, slope, length-width and elliptic fit were also evaluated. The results showed that at the time the LiDAR data was collected (June 24, 2014, the tree cover was around 25.11 % (or 15,674,341.8 m2 of the city’s urban barangays (or 62,426,064.6 m2. Among all urban barangays in Cebu City, Barangay Busay had the highest cover (55.79 % while barangay Suba had the lowest (0.8 %. The 16 barangays with less than 10 % tree cover were generally located in the coastal area, presumably due to accelerated urbanization. Thirty-one barangays have tree cover ranging from 10.59–-27.3 %. Only 3 barangays (i.e., Lahug, Talamban, and Busay have tree cover greater than 30 %. The overall accuracy of the analysis was 96.6 % with the Kappa Index of Agreement or KIA of 0.9. From the study, a grouping can be made of the city’s urban barangays with regards to tree cover. The grouping will be useful to urban planners not only in allocating budget to the tree planting program of the city but also in planning and creation of urban parks and playgrounds.

Automated Detection of Selective Logging in Amazon Forests Using Airborne Lidar Data and Pattern Recognition Algorithms

Science.gov (United States)

Keller, M. M.; d'Oliveira, M. N.; Takemura, C. M.; Vitoria, D.; Araujo, L. S.; Morton, D. C.

2012-12-01

Selective logging, the removal of several valuable timber trees per hectare, is an important land use in the Brazilian Amazon and may degrade forests through long term changes in structure, loss of forest carbon and species diversity. Similar to deforestation, the annual area affected by selected logging has declined significantly in the past decade. Nonetheless, this land use affects several thousand km2 per year in Brazil. We studied a 1000 ha area of the Antimary State Forest (FEA) in the State of Acre, Brazil (9.304 ○S, 68.281 ○W) that has a basal area of 22.5 m2 ha-1 and an above-ground biomass of 231 Mg ha-1. Logging intensity was low, approximately 10 to 15 m3 ha-1. We collected small-footprint airborne lidar data using an Optech ALTM 3100EA over the study area once each in 2010 and 2011. The study area contained both recent and older logging that used both conventional and technologically advanced logging techniques. Lidar return density averaged over 20 m-2 for both collection periods with estimated horizontal and vertical precision of 0.30 and 0.15 m. A relative density model comparing returns from 0 to 1 m elevation to returns in 1-5 m elevation range revealed the pattern of roads and skid trails. These patterns were confirmed by ground-based GPS survey. A GIS model of the road and skid network was built using lidar and ground data. We tested and compared two pattern recognition approaches used to automate logging detection. Both segmentation using commercial eCognition segmentation and a Frangi filter algorithm identified the road and skid trail network compared to the GIS model. We report on the effectiveness of these two techniques.

Automatic landslide detection from LiDAR DTM derivatives by geographic-object-based image analysis based on open-source software

Science.gov (United States)

Knevels, Raphael; Leopold, Philip; Petschko, Helene

2017-04-01

With high-resolution airborne Light Detection and Ranging (LiDAR) data more commonly available, many studies have been performed to facilitate the detailed information on the earth surface and to analyse its limitation. Specifically in the field of natural hazards, digital terrain models (DTM) have been used to map hazardous processes such as landslides mainly by visual interpretation of LiDAR DTM derivatives. However, new approaches are striving towards automatic detection of landslides to speed up the process of generating landslide inventories. These studies usually use a combination of optical imagery and terrain data, and are designed in commercial software packages such as ESRI ArcGIS, Definiens eCognition, or MathWorks MATLAB. The objective of this study was to investigate the potential of open-source software for automatic landslide detection based only on high-resolution LiDAR DTM derivatives in a study area within the federal state of Burgenland, Austria. The study area is very prone to landslides which have been mapped with different methodologies in recent years. The free development environment R was used to integrate open-source geographic information system (GIS) software, such as SAGA (System for Automated Geoscientific Analyses), GRASS (Geographic Resources Analysis Support System), or TauDEM (Terrain Analysis Using Digital Elevation Models). The implemented geographic-object-based image analysis (GEOBIA) consisted of (1) derivation of land surface parameters, such as slope, surface roughness, curvature, or flow direction, (2) finding optimal scale parameter by the use of an objective function, (3) multi-scale segmentation, (4) classification of landslide parts (main scarp, body, flanks) by k-mean thresholding, (5) assessment of the classification performance using a pre-existing landslide inventory, and (6) post-processing analysis for the further use in landslide inventories. The results of the developed open-source approach demonstrated good

Using GeoEye-1 Imagery for Multi-Temporal Object-Based Detection of Canegrub Damage in Sugarcane Fields in Queensland, Australia

KAUST Repository

Johansen, Kasper

2017-12-18

The greyback canegrub (Dermolepida albohirtum) is the main pest of sugarcane crops in all cane-growing regions between Mossman (16.5°S) and Sarina (21.5°S) in Queensland, Australia. In previous years, high infestations have cost the industry up to $40 million. However, identifying damage in the field is difficult due to the often impenetrable nature of the sugarcane crop. Satellite imagery offers a feasible means of achieving this by examining the visual characteristics of stool tipping, changed leaf color, and exposure of soil in damaged areas. The objective of this study was to use geographic object-based image analysis (GEOBIA) and high-spatial resolution GeoEye-1 satellite imagery for three years to map canegrub damage and develop two mapping approaches suitable for risk mapping. The GEOBIA mapping approach for canegrub damage detection was evaluated over three selected study sites in Queensland, covering a total of 254 km2 and included five main steps developed in the eCognition Developer software. These included: (1) initial segmentation of sugarcane block boundaries; (2) classification and subsequent omission of fallow/harvested fields, tracks, and other non-sugarcane features within the block boundaries; (3) identification of likely canegrub-damaged areas with low NDVI values and high levels of image texture within each block; (4) the further refining of canegrub damaged areas to low, medium, and high likelihood; and (5) risk classification. The validation based on field observations of canegrub damage at the time of the satellite image capture yielded producer’s accuracies between 75% and 98.7%, depending on the study site. Error of commission occurred in some cases due to sprawling, drainage issues, wind, weed, and pig damage. The two developed risk mapping approaches were based on the results of the canegrub damage detection. This research will improve decision making by growers affected by canegrub damage.

A study of earthquake-induced building detection by object oriented classification approach

Science.gov (United States)

Sabuncu, Asli; Damla Uca Avci, Zehra; Sunar, Filiz

2017-04-01

Among the natural hazards, earthquakes are the most destructive disasters and cause huge loss of lives, heavily infrastructure damages and great financial losses every year all around the world. According to the statistics about the earthquakes, more than a million earthquakes occur which is equal to two earthquakes per minute in the world. Natural disasters have brought more than 780.000 deaths approximately % 60 of all mortality is due to the earthquakes after 2001. A great earthquake took place at 38.75 N 43.36 E in the eastern part of Turkey in Van Province on On October 23th, 2011. 604 people died and about 4000 buildings seriously damaged and collapsed after this earthquake. In recent years, the use of object oriented classification approach based on different object features, such as spectral, textural, shape and spatial information, has gained importance and became widespread for the classification of high-resolution satellite images and orthophotos. The motivation of this study is to detect the collapsed buildings and debris areas after the earthquake by using very high-resolution satellite images and orthophotos with the object oriented classification and also see how well remote sensing technology was carried out in determining the collapsed buildings. In this study, two different land surfaces were selected as homogenous and heterogeneous case study areas. In the first step of application, multi-resolution segmentation was applied and optimum parameters were selected to obtain the objects in each area after testing different color/shape and compactness/smoothness values. In the next step, two different classification approaches, namely "supervised" and "unsupervised" approaches were applied and their classification performances were compared. Object-based Image Analysis (OBIA) was performed using e-Cognition software.

Object-based Forest Fire Analysis for Pedrógão Grande Fire Using Landsat 8 OLI and Sentinel-2A Imagery

Science.gov (United States)

Tonbul, H.; Kavzoglu, T.

2017-12-01

Forest fires are among the most important natural disasters with the damage to the natural habitat and human-life. Mapping damaged forest fires is crucial for assessing ecological effects caused by fire, monitoring land cover changes and modeling atmospheric and climatic effects of fire. In this context, satellite data provides a great advantage to users by providing a rapid process of detecting burning areas and determining the severity of fire damage. Especially, Mediterranean ecosystems countries sets the suitable conditions for the forest fires. In this study, the determination of burnt areas of forest fire in Pedrógão Grande region of Portugal occurred in June 2017 was carried out using Landsat 8 OLI and Sentinel-2A satellite images. The Pedrógão Grande fire was one of the largest fires in Portugal, more than 60 people was killed and thousands of hectares were ravaged. In this study, four pairs of pre-fire and post-fire top of atmosphere (TOA) and atmospherically corrected images were utilized. The red and near infrared (NIR) spectral bands of pre-fire and post-fire images were stacked and multiresolution segmentation algorithm was applied. In the segmentation processes, the image objects were generated with estimated optimum homogeneity criteria. Using eCognition software, rule sets have been created to distinguish unburned areas from burned areas. In constructing the rule sets, NDVI threshold values were determined pre- and post-fire and areas where vegetation loss was detected using the NDVI difference image. The results showed that both satellite images yielded successful results for burned area discrimination with a very high degree of consistency in terms of spatial overlap and total burned area (over 93%). Object based image analysis (OBIA) was found highly effective in delineation of burnt areas.

Global Warming and Geographically Scalar Climatic Objects Exist: An Ontologically Realist and Object-Oriented Analysis of the Daymet TMAX Climate Summaries for North America

Science.gov (United States)

Jackson, C. P.

2017-12-01

The scientific materialist worldview, what Peter Unger refers to as the Scientiphical worldview, or Scientiphicalism, has been utterly catastrophic for mesoscale objects in general, but, with its closely associated twentieth-century formal logic, this has been especially true for notoriously vague things like climate change, coastlines, mountains and dust storms. That is, any so-called representations or references ultimately suffer the same ontological demise as their referents, no matter how well-defined their boundaries may in fact be. Against this reductionist metaphysics, climatic objects are discretized within three separate ontologically realist systems, Graham Harman's object-oriented philosophy, or ontology (OOO), Markus Gabriel's ontology of fields of sense (OFS) and Tristan Garcia's two systems and new order of time, so as to make an ontological case for any geographically scalar object, beginning with pixels, as well as any notoriously vague thing they are said to represent. Four-month overlapping TMAX seasonals were first developed from the Oak Ridge National Laboratory (ORNL) Daymet climate temperature maximum (TMAX) monthly summaries (1980-2016) for North America and segmented within Trimble's eCognition Developer using the simple and widely familiar quadtree algorithm with a scale parameter of four, in this example. The regression coefficient was then calculated for the resulting 37-year climatic objects and an equally simple classification was applied. The same segmentation and classification was applied to the Daymet annual summaries, as well, for comparison. As was expected, the mean warming and cooling trends are lowest for the annual summary TMAX climatic objects. However, the Fall (SOND) season has the largest and smallest areas of warming and cooling, respectively, and the highest mean trend for warming objects. Conversely, Spring (MAMJ) has the largest and smallest areas undergoing cooling and warming, respectively. Finally, Summer (JJAS

High resolution remote sensing of water surface patterns

Science.gov (United States)

Woodget, A.; Visser, F.; Maddock, I.; Carbonneau, P.

2012-12-01

spectral and textural images properties is tested using Definiens eCognition software. Where possible, the DEM of the water surface topography is also analysed for identifying SFTs. The site is revisited in order to assess the temporal variability of SFTs as relating to changes in flow level, and the potential for variability in identifying SFTs from imagery as relating to changes in lighting and weather conditions.

High spatial resolution mapping of land cover types in a priority area for conservation in the Brazilian savanna

Science.gov (United States)

Ribeiro, F.; Roberts, D. A.; Hess, L. L.; Davis, F. W.; Caylor, K. K.; Nackoney, J.; Antunes Daldegan, G.

2017-12-01

-classification modification was performed to refine information to the major physiognomic groups of each ecosystem type. In this step, we used segmentation in eCognition, considering the random forest classification as input as well as other environmental layers (e.g. slope, soil types), which improved the overall classification to 75%.

Mapping of invasive Acacia species in Brazilian Mussununga ecosystems using high- resolution IR remote sensing data acquired with an autonomous Unmanned Aerial System (UAS)

Science.gov (United States)

Lehmann, Jan Rudolf Karl; Zvara, Ondrej; Prinz, Torsten

2015-04-01

The biological invasion of Australian Acacia species in natural ecosystems outside Australia has often a negative impact on native and endemic plant species and the related biodiversity. In Brazil, the Atlantic rainforest of Bahia and Espirito Santo forms an associated type of ecosystem, the Mussununga. In our days this biologically diverse ecosystem is negatively affected by the invasion of Acacia mangium and Acacia auriculiformis, both introduced to Brazil by the agroforestry to increase the production of pulp and high grade woods. In order to detect the distribution of Acacia species and to monitor the expansion of this invasion the use of high-resolution imagery data acquired with an autonomous Unmanned Aerial System (UAS) proved to be a very promising approach. In this study, two types of datasets - CIR and RGB - were collected since both types provide different information. In case of CIR imagery attention was paid on spectral signatures related to plants, whereas in case of RGB imagery the focus was on surface characteristics. Orthophoto-mosaics and DSM/DTM for both dataset were extracted. RGB/IHS transformations of the imagery's colour space were utilized, as well as NDVIblue index in case of CIR imagery to discriminate plant associations. Next, two test areas were defined in order validate OBIA rule sets using eCognition software. In case of RGB dataset, a rule set based on elevation distinction between high vegetation (including Acacia) and low vegetation (including soils) was developed. High vegetation was classified using Nearest Neighbour algorithm while working with the CIR dataset. The IHS information was used to mask shadows, soils and low vegetation. Further Nearest Neighbour classification was used for distinction between Acacia and other high vegetation types. Finally an accuracy assessment was performed using a confusion matrix. One can state that the IHS information appeared to be helpful in Acacia detection while the surface elevation

UAV-Based Thermal Imaging for High-Throughput Field Phenotyping of Black Poplar Response to Drought

Directory of Open Access Journals (Sweden)

Riccardo Ludovisi

2017-09-01

Full Text Available Poplars are fast-growing, high-yielding forest tree species, whose cultivation as second-generation biofuel crops is of increasing interest and can efficiently meet emission reduction goals. Yet, breeding elite poplar trees for drought resistance remains a major challenge. Worldwide breeding programs are largely focused on intra/interspecific hybridization, whereby Populus nigra L. is a fundamental parental pool. While high-throughput genotyping has resulted in unprecedented capabilities to rapidly decode complex genetic architecture of plant stress resistance, linking genomics to phenomics is hindered by technically challenging phenotyping. Relying on unmanned aerial vehicle (UAV-based remote sensing and imaging techniques, high-throughput field phenotyping (HTFP aims at enabling highly precise and efficient, non-destructive screening of genotype performance in large populations. To efficiently support forest-tree breeding programs, ground-truthing observations should be complemented with standardized HTFP. In this study, we develop a high-resolution (leaf level HTFP approach to investigate the response to drought of a full-sib F2 partially inbred population (termed here ‘POP6’, whose F1 was obtained from an intraspecific P. nigra controlled cross between genotypes with highly divergent phenotypes. We assessed the effects of two water treatments (well-watered and moderate drought on a population of 4603 trees (503 genotypes hosted in two adjacent experimental plots (1.67 ha by conducting low-elevation (25 m flights with an aerial drone and capturing 7836 thermal infrared (TIR images. TIR images were undistorted, georeferenced, and orthorectified to obtain radiometric mosaics. Canopy temperature (Tc was extracted using two independent semi-automated segmentation techniques, eCognition- and Matlab-based, to avoid the mixed-pixel problem. Overall, results showed that the UAV platform-based thermal imaging enables to effectively assess genotype

Decision Tree Repository and Rule Set Based Mingjiang River Estuarine Wetlands Classifaction

Science.gov (United States)

Zhang, W.; Li, X.; Xiao, W.

2018-05-01

The increasing urbanization and industrialization have led to wetland losses in estuarine area of Mingjiang River over past three decades. There has been increasing attention given to produce wetland inventories using remote sensing and GIS technology. Due to inconsistency training site and training sample, traditionally pixel-based image classification methods can't achieve a comparable result within different organizations. Meanwhile, object-oriented image classification technique shows grate potential to solve this problem and Landsat moderate resolution remote sensing images are widely used to fulfill this requirement. Firstly, the standardized atmospheric correct, spectrally high fidelity texture feature enhancement was conducted before implementing the object-oriented wetland classification method in eCognition. Secondly, we performed the multi-scale segmentation procedure, taking the scale, hue, shape, compactness and smoothness of the image into account to get the appropriate parameters, using the top and down region merge algorithm from single pixel level, the optimal texture segmentation scale for different types of features is confirmed. Then, the segmented object is used as the classification unit to calculate the spectral information such as Mean value, Maximum value, Minimum value, Brightness value and the Normalized value. The Area, length, Tightness and the Shape rule of the image object Spatial features and texture features such as Mean, Variance and Entropy of image objects are used as classification features of training samples. Based on the reference images and the sampling points of on-the-spot investigation, typical training samples are selected uniformly and randomly for each type of ground objects. The spectral, texture and spatial characteristics of each type of feature in each feature layer corresponding to the range of values are used to create the decision tree repository. Finally, with the help of high resolution reference images, the

The Urban Heat Island Impact in Consideration of Spatial Pattern of Urban Landscape and Structure

Science.gov (United States)

Kim, J.; Lee, D. K.; Jeong, W.; Sung, S.; Park, J.

2015-12-01

Preceding study has established a clear relationship between land surface temperature and area of land covers. However, only few studies have specifically examined the effects of spatial patterns of land covers and urban structure. To examine how much the local climate is affected by the spatial pattern in highly urbanized city, we investigated the correlation between land surface temperature and spatial patterns of land covers. In the analysis of correlation, we categorized urban structure to four different land uses: Apartment residential area, low rise residential area, industrial area and central business district. Through this study, we aims to examine the types of residential structure and land cover pattern for reducing urban heat island and sustainable development. Based on land surface temperature, we investigated the phenomenon of urban heat island through using the data of remote sensing. This study focused on Daegu in Korea. This city, one of the hottest city in Korea has basin form. We used high-resolution land cover data and land surface temperature by using Landsat8 satellite image to examine 100 randomly selected sample sites of 884.15km2 (1)In each land use, we quantified several landscape-levels and class-level landscape metrics for the sample study sites. (2)In addition, we measured the land surface temperature in 3 year hot summer seasons (July to September). Then, we investigated the pattern of land surface temperature for each land use through Ecognition package. (3)We deducted the Pearson correlation coefficients between land surface temperature and each landscape metrics. (4)We analyzed the variance among the four land uses. (5)Using linear regression, we determined land surface temperature model for each land use. (6)Through this analysis, we aims to examine the best pattern of land cover and artificial structure for reducing urban heat island effect in highly urbanized city. The results of linear regression showed that proportional land

Mapping Depositional Facies on Great Bahama Bank: An Integration of Groundtruthing and Remote Sensing Methods

Science.gov (United States)

Hariss, M.; Purkis, S.; Ellis, J. M.; Swart, P. K.; Reijmer, J.

2013-12-01

Great Bahama Bank (GBB) has been used in many models to illustrate depositional facies variation across flat-topped, isolated carbonate platforms. Such models have served as subsurface analogs at a variety of scales. In this presentation we have integrated Landsat TM imagery, a refined bathymetric digital elevation model, and seafloor sample data compiled into ArcGIS and analyzed with eCognition to develop a depositional facies map that is more robust than previous versions. For the portion of the GBB lying to the west of Andros Island, the facies map was generated by pairing an extensive set of GPS-constrained field observations and samples (n=275) (Reijmer et al., 2009, IAS Spec Pub 41) with computer and manual interpretation of the Landsat imagery. For the remainder of the platform, which lacked such rigorous ground-control, the Landsat imagery was segmented into lithotopes - interpreted to be distinct bodies of uniform sediment - using a combination of edge detection, spectral and textural analysis, and manual editing. A map was then developed by assigning lithotopes to facies classes on the basis of lessons derived from the portion of the platform for which we had rigorous conditioning. The new analysis reveals that GBB is essentially a very grainy platform with muddier accumulations only in the lee of substantial island barriers; in this regard Andros Island, which is the largest island on GBB, exerts a direct control over the muddiest portion of GBB. Mudstones, wackestones, and mud-rich packstones cover 7%, 6%, and 15%, respectively, of the GBB platform top. By contrast, mud-poor packstones, grainstones, and rudstones account for 19%, 44%, and 3%, respectively. Of the 44% of the platform-top classified as grainstone, only 4% is composed of 'high-energy' deposits characterized by the development of sandbar complexes. The diversity and size of facies bodies is broadly the same on the eastern and western limb of the GBB platform, though the narrower eastern

UAV-Based Thermal Imaging for High-Throughput Field Phenotyping of Black Poplar Response to Drought.

Science.gov (United States)

Ludovisi, Riccardo; Tauro, Flavia; Salvati, Riccardo; Khoury, Sacha; Mugnozza Scarascia, Giuseppe; Harfouche, Antoine

2017-01-01

Poplars are fast-growing, high-yielding forest tree species, whose cultivation as second-generation biofuel crops is of increasing interest and can efficiently meet emission reduction goals. Yet, breeding elite poplar trees for drought resistance remains a major challenge. Worldwide breeding programs are largely focused on intra/interspecific hybridization, whereby Populus nigra L. is a fundamental parental pool. While high-throughput genotyping has resulted in unprecedented capabilities to rapidly decode complex genetic architecture of plant stress resistance, linking genomics to phenomics is hindered by technically challenging phenotyping. Relying on unmanned aerial vehicle (UAV)-based remote sensing and imaging techniques, high-throughput field phenotyping (HTFP) aims at enabling highly precise and efficient, non-destructive screening of genotype performance in large populations. To efficiently support forest-tree breeding programs, ground-truthing observations should be complemented with standardized HTFP. In this study, we develop a high-resolution (leaf level) HTFP approach to investigate the response to drought of a full-sib F 2 partially inbred population (termed here 'POP6'), whose F 1 was obtained from an intraspecific P. nigra controlled cross between genotypes with highly divergent phenotypes. We assessed the effects of two water treatments (well-watered and moderate drought) on a population of 4603 trees (503 genotypes) hosted in two adjacent experimental plots (1.67 ha) by conducting low-elevation (25 m) flights with an aerial drone and capturing 7836 thermal infrared (TIR) images. TIR images were undistorted, georeferenced, and orthorectified to obtain radiometric mosaics. Canopy temperature ( T c ) was extracted using two independent semi-automated segmentation techniques, eCognition- and Matlab-based, to avoid the mixed-pixel problem. Overall, results showed that the UAV platform-based thermal imaging enables to effectively assess genotype

Integrating Landsat Data and High-Resolution Imagery for Applied Conservation Assessment of Forest Cover in Latin American Heterogenous Landscapes

Science.gov (United States)

Thomas, N.; Rueda, X.; Lambin, E.; Mendenhall, C. D.

2012-12-01

Large intact forested regions of the world are known to be critical to maintaining Earth's climate, ecosystem health, and human livelihoods. Remote sensing has been successfully implemented as a tool to monitor forest cover and landscape dynamics over broad regions. Much of this work has been done using coarse resolution sensors such as AVHRR and MODIS in combination with moderate resolution sensors, particularly Landsat. Finer scale analysis of heterogeneous and fragmented landscapes is commonly performed with medium resolution data and has had varying success depending on many factors including the level of fragmentation, variability of land cover types, patch size, and image availability. Fine scale tree cover in mixed agricultural areas can have a major impact on biodiversity and ecosystem sustainability but may often be inadequately captured with the global to regional (coarse resolution and moderate resolution) satellite sensors and processing techniques widely used to detect land use and land cover changes. This study investigates whether advanced remote sensing methods are able to assess and monitor percent tree canopy cover in spatially complex human-dominated agricultural landscapes that prove challenging for traditional mapping techniques. Our study areas are in high altitude, mixed agricultural coffee-growing regions in Costa Rica and the Colombian Andes. We applied Random Forests regression tree analysis to Landsat data along with additional spectral, environmental, and spatial variables to predict percent tree canopy cover at 30m resolution. Image object-based texture, shape, and neighborhood metrics were generated at the Landsat scale using eCognition and included in the variable suite. Training and validation data was generated using high resolution imagery from digital aerial photography at 1m to 2.5 m resolution. Our results are promising with Pearson's correlation coefficients between observed and predicted percent tree canopy cover of .86 (Costa

A comparison of the accuracy of pixel based and object based classifications of integrated optical and LiDAR data

Science.gov (United States)

Gajda, Agnieszka; Wójtowicz-Nowakowska, Anna

2013-04-01

A comparison of the accuracy of pixel based and object based classifications of integrated optical and LiDAR data Land cover maps are generally produced on the basis of high resolution imagery. Recently, LiDAR (Light Detection and Ranging) data have been brought into use in diverse applications including land cover mapping. In this study we attempted to assess the accuracy of land cover classification using both high resolution aerial imagery and LiDAR data (airborne laser scanning, ALS), testing two classification approaches: a pixel-based classification and object-oriented image analysis (OBIA). The study was conducted on three test areas (3 km2 each) in the administrative area of Kraków, Poland, along the course of the Vistula River. They represent three different dominating land cover types of the Vistula River valley. Test site 1 had a semi-natural vegetation, with riparian forests and shrubs, test site 2 represented a densely built-up area, and test site 3 was an industrial site. Point clouds from ALS and ortophotomaps were both captured in November 2007. Point cloud density was on average 16 pt/m2 and it contained additional information about intensity and encoded RGB values. Ortophotomaps had a spatial resolution of 10 cm. From point clouds two raster maps were generated: intensity (1) and (2) normalised Digital Surface Model (nDSM), both with the spatial resolution of 50 cm. To classify the aerial data, a supervised classification approach was selected. Pixel based classification was carried out in ERDAS Imagine software. Ortophotomaps and intensity and nDSM rasters were used in classification. 15 homogenous training areas representing each cover class were chosen. Classified pixels were clumped to avoid salt and pepper effect. Object oriented image object classification was carried out in eCognition software, which implements both the optical and ALS data. Elevation layers (intensity, firs/last reflection, etc.) were used at segmentation stage due to

Dynamic Post-Earthquake Image Segmentation with an Adaptive Spectral-Spatial Descriptor

Directory of Open Access Journals (Sweden)

Genyun Sun

2017-08-01

, including the fractal net evolution approach (FNEA, as implemented in the eCognition software, Trimble Inc., Westminster, CO, USA, the J-value segmentation (JSEG method, the graph-based segmentation (GSEG method, and the statistical region merging (SRM approach. The experiments were conducted on six VHR subarea images captured by RGB sensors mounted on aerial platforms, which were acquired after the 2008 Wenchuan Ms 8.0 earthquake. Quantitative and qualitative assessments demonstrated that the proposed method offers high feasibility and improved accuracy in the segmentation of post-earthquake VHR aerial images.

Object-based landslide detection in different geographic regions

Science.gov (United States)

Friedl, Barbara; Hölbling, Daniel; Eisank, Clemens; Blaschke, Thomas

2015-04-01

, SPOT-5 images are combined with digital elevation models (DEM) for developing a consistent semi-automated landslide detection approach using eCognition (Trimble) software. Suitable image objects are generated by means of multiresolution segmentation. Expert knowledge, i.e. reported facts on features (e.g. mean object slope, mean NDVI) and thresholds that are commonly chosen by professionals for digital landslide mapping, is considered during classification. The applicability of a range of features is tested and the most promising parameters, i.e. features that produce appropriate results for both regions, are selected for landslide detection. However, minor adaptations of particular thresholds are necessary due to the distinct environmental conditions of the test sites. In order to reduce the number of required adjustments to a minimum, relational features and spectral indices are primarily used for classification. The obtained results are finally compared to manually digitized reference polygons and existing landslide inventories in order to quantify the applicability of the developed object-based landslide detection approach in different geographic regions.

Combining TerraSAR-X and SPOT-5 data for object-based landslide detection

Science.gov (United States)

Friedl, B.; Hölbling, D.; Füreder, P.

2012-04-01

Landslide detection and classification is an essential requirement in pre- and post-disaster hazard analysis. In earlier studies landslide detection often was achieved through time-consuming and cost-intensive field surveys and visual orthophoto interpretation. Recent studies show that Earth Observation (EO) data offer new opportunities for fast, reliable and accurate landslide detection and classification, which may conduce to an effective landslide monitoring and landslide hazard management. To ensure the fast recognition and classification of landslides at a regional scale, a (semi-)automated object-based landslide detection approach is established for a study site situated in the Huaguoshan catchment, Southern Taiwan. The study site exhibits a high vulnerability to landslides and debris flows, which are predominantly typhoon-induced. Through the integration of optical satellite data (SPOT-5 with 2.5 m GSD), SAR (Synthetic Aperture Radar) data (TerraSAR-X Spotlight with 2.95 m GSD) and digital elevation information (DEM with 5 m GSD) including its derived products (e.g. slope, curvature, flow accumulation) landslides may be examined in a more efficient way as if relying on single data sources only. The combination of optical and SAR data in an object-based image analysis (OBIA) domain for landslide detection and classification has not been investigated so far, even if SAR imagery show valuable properties for landslide detection, which differ from optical data (e.g. high sensitivity to surface roughness and soil moisture). The main purpose of this study is to recognize and analyze existing landslides by applying object-based image analysis making use of eCognition software. OBIA provides a framework for examining features defined by spectral, spatial, textural, contextual as well as hierarchical properties. Objects are derived through image segmentation and serve as input for the classification process, which relies on transparent rulesets, representing knowledge

Vegetation classification and distribution mapping report Mesa Verde National Park

Science.gov (United States)

Thomas, Kathryn A.; McTeague, Monica L.; Ogden, Lindsay; Floyd, M. Lisa; Schulz, Keith; Friesen, Beverly A.; Fancher, Tammy; Waltermire, Robert G.; Cully, Anne

2009-01-01

The classification and distribution mapping of the vegetation of Mesa Verde National Park (MEVE) and surrounding environment was achieved through a multi-agency effort between 2004 and 2007. The National Park Service’s Southern Colorado Plateau Network facilitated the team that conducted the work, which comprised the U.S. Geological Survey’s Southwest Biological Science Center, Fort Collins Research Center, and Rocky Mountain Geographic Science Center; Northern Arizona University; Prescott College; and NatureServe. The project team described 47 plant communities for MEVE, 34 of which were described from quantitative classification based on f eld-relevé data collected in 1993 and 2004. The team derived 13 additional plant communities from field observations during the photointerpretation phase of the project. The National Vegetation Classification Standard served as a framework for classifying these plant communities to the alliance and association level. Eleven of the 47 plant communities were classified as “park specials;” that is, plant communities with insufficient data to describe them as new alliances or associations. The project team also developed a spatial vegetation map database representing MEVE, with three different map-class schemas: base, group, and management map classes. The base map classes represent the fi nest level of spatial detail. Initial polygons were developed using Definiens Professional (at the time of our use, this software was called eCognition), assisted by interpretation of 1:12,000 true-color digital orthophoto quarter quadrangles (DOQQs). These polygons (base map classes) were labeled using manual photo interpretation of the DOQQs and 1:12,000 true-color aerial photography. Field visits verified interpretation concepts. The vegetation map database includes 46 base map classes, which consist of associations, alliances, and park specials classified with quantitative analysis, additional associations and park specials noted