ID: 57397
Title: Registration of multisource satellite images by thin-plate splines with highly reliable conjugate points
Author: Joz Wu, Chi Chang, Hsien-Yu Tsai, and Ming-Che Liu
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 6, June 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: Image registration, algorithm
Abstract: Image registration is essential for geospatial information systems anlaysis, which usually involves integrating multitemporal and multispectral datasets from remote optical and radar sensors. An algorithm that deals with feature extraction, keypoint matching, outlier detection, and image warping is experimental in this study. The methods currently available in the literature rely on techniques, such as the scale-invariant feature transform, between-edge cost minimization, normalized cross correlation, least-squares image matching, random sample consensus, iterated data snooping, and thin-plate splines. Their basics are high-lighted and encoded into a computer program. The test images are excerpts from digital files created by the multispectral SPOT-5 and Formosat -2 sensors, side-looking ERS-2, and Envisat synthetic aperture radars, and by the panchromatic Ikonos and QuickBird sensors. Suburban areas, housing rooftops, the countryside and hilly plantations are studied. The co-registered images are displayed with block sub-images in a criss-cross pattern. Besides the imagery, the registration accuracy is expressed by the root mean square error. Toward the end, this paper also includes a few opinions on issues that are believed to hinder a correct correspondence between diverse images.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57396
Title: A comparison of SRTM V4 and ASTER GDEM for hydrological applications in low relief terrain
Author: Mariano Moreno-de las Heras, Patricia M Saco, and Garry R Willgoose
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 7, July 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: Shuttle Radar Topography Mission elevation model (SRTM V4), ASTER GDEM
Abstract: We compare the performance of the latest version of the Shuttle Radar Topography Mission elevation model (SRTM V4) with the ASTER-derived model (ASTER GDEM Version 1) for the determination of hydrological and geomorphological descriptors in low gradient Australian landscapes. The vertical quantizatin of these models (1 m) limited the landform representation, generating flat areas that required extensive preprocessing to produce hydrologically connected surfaces. The ASTER GDEM was more affected by surface filtering (i.e., depression filling and flat areas treatment), especially in areas containing systematic artifacts (eg., pits, steps) that changed network properties. The vertical data accuracy of these models failed to resolve uncertainties associated with flow routing in nearly flat areas. We conclude that the SRTM V4 is a more reliable model, particularly in areas where the ASTER GDEM displays elevation artifacts. However, its performance is constrained due to the lack of both adequate data accuracy, and sub-meter vertical detail.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57395
Title: Application of time series Landsat images to examining land-use/land-cover dynamic change
Author: Dengsheng Lu, Scott Hetrick, Emilio Moran, and Guiying Li
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 7, July 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: Land use, land cover
Abstract: A hierarchial - based classification method was designed to develop time series land-use/land -cover datasets from Landsat images between 1977 and 2008 in Lucas do Rio Verde, Mato Grosso, Brazil. A post-classification comparison approach was used to examine land-use/land-cover change trajectories, which emphasis is on the conversions from vegetation or agropasture to impervious surface area, from vegetation to agropasture, and from agropasture to regenerating vegetation. Results of this research indicated that increase in impervious surface area mainly resulted from the loss of cerrado in the initial decade of the study period and from loss of agricultural lands in the last two decades. Increase in agropasture was mainly at the expense of losing cerrado in the first two decades and relatively evenly from the loss of primary forest and cerrado in the last decade. When impervious surface area was less than approaximately 40 km2 before 1999, impervious surface area was negatively related to cerrado and forest, and positively related to agropasture areas, but after impervious surface area reached agropasture areas, but after impervioius surface area reached 40 km2 in 1999, no obvious relathioship exists between them.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57394
Title: Building detection in complex scenes thorough effective separation of buildings from trees
Author: Mohammad Awrangjeb, Chunsun Zhang, and Clive S Fraser
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 7, July 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: normalized DSM, trees, buildings
Abstract: Effective separation of buildings from trees is a major challenge in image-based automatic building detection. This paper presents a three-step method for effective separation of buildings from trees using aerial imagery and lidar data. First, it uses cues such as height to remove objects of low height such as bushes, and width to exclude trees with small horizontal coverage. The height threshold is also used to generate a ground mask where buildings are found to be more separately than in so-called normalized DSM. Second, image entropy and color information are jointly applied to remove easily distinguishable trees. Finally, an innovative rule-based procedure is employed using the edge orientation histogram from the imagery to eliminate false positive candidates. The improved building detection algorithm has been tested on different test areas and it is shown that the algorithm offers high building detection rate in complex scenes which are hilly and densely vegetated.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57393
Title: Modeling percent tree canopy cover: A pilot study
Author: John W Coulston, Gretchen G Moisen, Barry T Wilson, Mark V Finco, Warren B Cohen, and C Kenneth Brewer
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 7, July 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: canopy, modeling techniques, random forests and beta regression
Abstract: Tree canopy cover is a fundamental component of the landscape, and the amount of cover influences fire behavior, air pollution mitigation, and carbon storage. As such, efforts to empirically model percent tree canopy cover across the United States are a critical area of research. The 2001 national -scale canopy cover modeling and mapping effort was completed in 2006, and here we present results from a pilot study for a 2011 product. We examined the influence of two different modeling techniques (random forests and beta regression), two different Landsat imagery normalization processes, and eight different sampling intensities across five different pilot areas. We found that random forest out-performed beta regression techniques and that there was little difference between models developed based on the two different normalization techniques. Based on these results we present a prototype study design which will test canopy cover modeling approaches across a broader spatial scale.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57392
Title: Computing non-crossing smooth contours on triangulated meshes
Author: Yurai Nunez-Rodriguez, Michael A Johnson, Igor Raskin, and Jesse L The
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 7, July 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: Contour processing, Photogrammetry, Meteorology, Geographic Information Science
Abstract: Contour processing is a fundamental task in many fields, such as Photogrammetry, Meteorology, and Geographic Information Science. Contours, usually extracted from raster images or triangulated meshes, greatly benefit from a smoothing postprocessing phase. In the case of triangulated meshes, unsmoothed contours look angular, and disclose the edges of the underlying mesh. Attempts to smooth contours individually by using splines may cause crossing among nearby contours, clearly misrepresenting the data. To partially reduce this problem, previous studies propose to use the underlying mesh at aiding the contour smoothing process. The algorithm proposed herein builds on previous work to guarantee, for the first time, non-crossing smooth contours that are continuous in the first derivative. We formally prove our claims and present experimental results to show that, in addition, our algorithm outputs less than half the control points produced by previous approaches. Further work is proposed for achieving higher degrees of smoothness.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57391
Title: GeoEye imagery and Lidar technology for small-area population estimation: An epidemiological viewpoint
Author: Erika Upegui and Jean-Francois Viel
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 7, July 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: Population model, GeoEye imagery
Abstract: There is a continuing need for alternative approaches to obtain small-area populations when population census information is unavailable or inaccurate. The aim of this study was to develop an automated and exportable population model, based on GeoEye imagery and lidar technology, for estimating population counts and calculating disease rates in small areas. A satellite image (covering the City of Besancon, France) was processed to extract built-up pixels using ISODATA and hierarchial classifications. A dasymetric method was used to calculate population estimates at the block group level. Female breast cancer incidence rates were computed for each block group with the number of cases as numerator, and alternatively the population estimates and census data as denominators. A strong agreement was found between the estimated and the observed (census-based) incidence rates. This apportioning procedure could be of special interest to public health decision makers facing a lack of population census data.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57390
Title: A fusion approach for tree crown delineation from Lidar data
Author: Colin J Gleason and Jungho Im
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 7, July 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: COTH method, lidar data
Abstract: This study proposes a multi-step method (the COTH method) to delineate individual tree crowns in dense forest conditions using lidar data, with the intent for the final delineation results to be used in a biomass estimation study. The study was conducted for an even-aged Norway Spruce (Picea abies) plantation containing 188 trees located in Tully, New York, and owned by the State University of New York College of Environmental Science and Forestry (SUNY ESF). Lidar data with a point density of 12.7 points/m2 was collected in August 2010, and field data were collected to measure tree height and species in August 2010 as well. Field data containing tree height and crown width, an important component of treetop detection, were collected in summer 2006. By combining heuristicaly (genetic algorithm) opitmized object recognition to detect tree crown objects, local maxima filtering with variable window size to detect treetops, and a modified hill climbing algorithm to segment crown objects; treetops were identified with 86.2 percent accuracy and 23.9 percent commision error. The overall areal accuracy of the delineation was 72.5 percent. The automated COTH method represents an improvement in crown delineation accuracy for lidar data.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57389
Title: Three decades of war and food insecurity in Iraq
Author: Glen R Gibson, James B Campbell, Randolph H Wynne
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 8, August 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: Landsat TM imagery, ETM+ imagery,normalized difference vegetative index (NDVI)
Abstract: Iraq has suffered more than three decades of near continuous war and instability, affecting its food security. Our objectives were to assess whether cultivated area in central Iraq, as estimated using Landsat TM and ETM+ imagery, changed during and as a result of three decades of war and sanctions, and to determine which period of conflict experienced the greatest change. We created multitemporal features of maximum normalized difference vegetative index (NDVI) per pixel and used decision tree analysis to classify pixels with high maximum NDVI as cultivated. Validation with higher resolution imagery showed a 94 percent overall classificaiton accuracy and a kappa statistic of 0.88. The results indicated that cultivated area changed little between the Iran-Iraq War (1980 to 1988) and the Gulf War (1990 to 1991), increased by 20 percent (from 1.72 to 2.04 Mha) during the preiod of United Nations sanctions (1990 to 2003), and dropped to below pre-sanction levels during Operation Iraqi Freedom (2003 go 2011).
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57388
Title: Development of an advanced Global Field Survey System (GFSS) for terrestrial monitoring and mapping with a demonstration for agricultural cropland mapping in Asia
Author: An Ngoc Van, Kyaw Sann Oo, and Wataru Takeuchi
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 8, August 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: GFSS, GPS
Abstract: Field survey database always play a critical role in terrestrial monitoring and mapping. This study develops an advanced GFSS as a tool for conducting field studies and builds a field survey database. The GFSS allows users to use Apple ' s iPhone or iPad to collect field survey data, including photos, locations, and the user-defined data. The field survey data gathered with an iPhone or iPad are synchronized with a GPS Photo database on a server and can be shared with other users. Existing field survey data can be imported to the GPS Photo database and the data in the GPS Photo database can be extracted to various formats. As a demonstration, the GPS Photo database was utilized to assess the accuracy of the irrigated area maps of Vietnam, Laos, and Myanmar, which were derived from the Global Irrigated Area Map that was released by the International Water Management Institute in 2007.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57387
Title: Determining irrigated areas and quantifying blue water use in Europe using Remote sensing Meteosat Second Generation (MSG) products and Global Land Data Assimilation System (GLDAS) data
Author: Mireia Romaguera, Maarten S Krol, Mhd. Suhyb Salama, Arjen Y Hoekstra, and Zhongbo Su
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 8, August 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: Global Land Data Assimilation System (GLDAS) , Meteo-sat Second Generation (MSG) satellites
Abstract: In this paper, we propose on innovative method for identifying irrigated areas and quantifying the blue evapotranspiration (ETb), or irrigation water evapotranspired from the field. The method compares actual ET (ETactual), or crop water use, values from the Global Land Data Assimilation System (GLDAS) and remote sensing based ETactual estimates obtained from Meteo-sat Second Generation (MSG) satellites. Since GLDAS simulations do not account for extra water supply due to irrigation, it is expected that they underestimate ETactual during the cropping season in irrigated areas. However, remote sensing techniques based on the energy balance are able to observe the total ETactual. In order to isolate irrigation effects from other fluctuations that may lead to discrepancies between the different ETactual products, the bias between model simulations and remote sensing observations was estimated using reference targets of rainfed (non-irrigated) croplands on a daily basis in different areas across the study region (Europe). Analysis of the yearly values of ETb (irrigated area and volume obtained for croplands in Europe for 2008) showed that the method identified irrigation when yearly values were higher than 50 mm. The accuracy of the method was assessed by analyzing the spatial representativity of the calculated biases adn evaluating the daily ETb values obtained. The irrigated areas were compared with the results provided by Siebert et al. (2007) and Thenkabail et al. (2009b), obtaining a spatial match of 47 and 72 percent, respectively, with overestimation of irrigated area on a country scale. Additional evaluation with the ETb results of Mekonnen and Hoekstra (2011) showed 75 percent of overlap for + 50mm range. Finally, validation with in situ data on irrigation volumes proved the cogency of our method with less than 20 percent differences between derived and measured values.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57386
Title: Estimating irrigated agricultural water use through Landsat TM and a simplified surface energy balance modeling in the semi-arid environments of Arizona
Author: Shai Kapan and Soe W Myint
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 8, August 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: Evapotranspiration (ET), SEBAL and RESET models,
Abstract: Quantifying evapotranspiration (ET) is a key element for achieving better water management, especially in regions where agriculture is the main water consumer. A hybrid model combining the SEBAL and RESET models (S-RESET) was developed to effectively estimate actual ET (water use) of the agriculture sector around the Phoenix metropolitan area. To examine how irrigated agriculture water consumption varies with climate, the S-RESET model was applied under wet and dry climatic conditions. Results show that the average ET for active agriculture is 9.3 mm/day (+ 3.8 mm/day) during the study period. Seasonal water use was 438 mm for 2000 (drought) and 494 mm for 2008 (wet). Based on the seasonal ET, we concluded that farmers in arid region use the same amount of water regardless of climatic conditions, implying that the agriculture sector as a whole may not be sensitive to drought as long as there is sufficient water from irrigation. This finding carries significant implications for the region ' s water security.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57385
Title: Mapping fragmented agricultural systems in the Sudano-Sahelian environments of Africa using random forest and ensemble metrics of coarse resolution MODIS imagery
Author: Elodie Vintrou, Mamy Soumare, Simon Bernard, Agnes Begue, Christian Baron, and Danny Lo Seen
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 8, August 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: Random Forest classification model, NDVI,
Abstract: We worked on the assumption that agricultural systems shaped the landscape through human cropping practices, and that the resulting landscape can be described with a set of coarse resolution satellite-derived metrics (spectral, textural, temporal, and spatial metrics). A Random Forest classification model was developed at the village scale in South Mali, based on 100 samples, with data on the main type of agricultural system in each village (three-class typology), and 30 MODIS - derived and socio-environmental metrics calculated on agricultural areas. The model was found to perform well (overall accuracy of 60 percent) and was stable. Class A (food crops) and B (intensive agriculture) displayed good producer ' s accuracy (70 percent and 67 percent, respectively), while class C (mixed agriculture) was less accurate (50 percent). The most important metrics were shown to be the annual mean of NDVI, followed by the phenology transition daes and texture metrics. However, when considering each set of metrics separately, texture emerged as the most discriminating factor (with 53 percent of global accuracy). This result, i.e. that even coarse resolution imagery contains textural information that can be used for crop mapping, is new. Such maps could be used for crop mapping, is new. Such maps could be used in food security systems as an indicator of system vulnerability, or as spatial inputs for crop yield models.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57384
Title: A shape-matching cropping index (CI) mapping method to determine agricultural cropland intensities in China using MODIS time-series data
Author: Jianhong Liu, Wenquan Zhu and Xuefeng Cui
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 8, August 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: Cropping Index (CI) , Vegetation Index (VI), Enhanced Vegetation Index (EVI),Moderate Resolution Imaging Spectroradiometer (MODIS)
Abstract: As an important indicator of cropping intensity, Cropping Index (CI) is defined as the number of peaks in the Vegetation Index (VI) time-series curve in a year. The existing CI mapping algorithms (e.g., the cross-fitting and the second order difference algorithm) are vulnerable to noise contained in the VI time series and need a priori knowledge and some extra constraints which could not be directly derived from the VI time-series data. In this paper, a shape-matching method is developed which can map CI directly from the preprocessed VI time-series data without the de-noising processes. This shape-matching method utilizes a decision -making process to find out the true peaks in the VI time-series curve based on a rank order mathematical morphology algorithm. The processing procedure involves five steps: (a) determination fo the temporal moving window size, (b) detection of local maximum/minimum points, (c) exclusion of false maximum/minimum points, (d) determinaiton of the threshold for the minimum points, (d) determination of the threshol for the minimum growth amplitude, and (e) mapping of Cl. This shape-matching method only needs two input parameters, the temporal moving window size and the threshold for the minimum growth amplitude, which can be both directly derived from the VI time-series data with some selected test pixels. Moreover, the response of the shape-matching method is relatively insensitive to the exact values of its design parameters, making it more flexible and effective in adapting to other regions. This new method is applied to map the CIs in Jiangsu Province, China, in 2010, based on the Enhanced Vegetation Index (EVI) time-series data derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) product. The overal CI mapping accuracy for the shape-matching method is 80 percent, which is much higher than the CI mapping accuracy of 60 percent for the second order difference algorithm. This shape-matching method can be further applied to other regions with a grid-search for its optimal parameters using some test pixels.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None
ID: 57383
Title: Mapping crop types, irrigated areas, and cropping intensities in heterogeneous landscapes of Southern India using multi-temporal medium-resolution imagery: implications for assessing water use in agriculture
Author: Elizabeth Heller, Jeanine M Rhemtulla, Sharachchandra Lele, Margaret Kalacska, Shrinivas Badiger, Raja Sengupta, and Navin Ramankutty
Editor: Russell G Congalton
Year: 2012
Publisher: ASPRS, Vol 78, No 8, August 2012
Source: Centre for Ecological Sciences
Reference: None
Subject: Photogrammetric Engineering and Remote Sensing
Keywords: irrigated land, crop land, multi- temporal imagery
Abstract: In regions of water scarcity, mapping individual crops, cropping intensities and irrigation can contribute significantly to understanding agricultural water use. But such mapping is challenging in landscapes dominated by small-scale traditional agricultural land holdings with high spatial and temporal heterogeneity. Here, we assessed the benefit of using multi-temporal 24 m resolution LISS-III imagery to characterize cropping systems in the Malaprabha basin of southern India. We used hierarchial stacked supervised cassification to create three increasingly detailed maps showing: (a) single rainfed paddy rice versus continuously irrigated sugarcane , (b) irrigated versus rainfed areas , and (C) multiple cropping. Although increasing detail was accompanied by decreasing overall accuracies (89 percent, 74.6 percent and 60.1 percent respectively), using multi- temporal imagery out- performed single imagery alone in all cases. Results also led to higher estimates of total (69.8 percent) and irrigated (34.7 percent) cropland than previous single - imagery studies and census data, revealing the high uncertainty in crop estimates in this region.
Location: TE 12, Biologicalsciences Building, IISc
Literature cited 1: None
Literature cited 2: None