ASPRS 2014 Annual Conference & co-located JACIE Workshop

  • Increase font size
  • Default font size
  • Decrease font size


ASPRS 2014 Annual Conference

 Geospatial Power in Our Pockets

& co-located JACIE Workshop
Joint Agency Commercial Imagery Evaluation (JACIE) Workshop

Louisville, Kentucky USA   *  March 23-28, 2014  *  The Galt House Hotel

Join the American Society for Photogrammetry and Remote Sensing(ASPRS) for the 2014 Annual Conference as we head to the home of the Kentucky Derby, the Louisville Slugger baseball bat and Southern Hospitality, Louisville, Kentucky, March 23 - 27, 2014!

This year we are excited to welcome the JACIE Workshop to co-locate in Louisville. The JACIE Workshop will be held March 26 - 28, 2014 at the Galt House Hotel and will be combining a general session and special technical sessions throughout the week with the ASPRS Conference. This is an exciting partnership for both organizations!


The intent of the JACIE workshop is to exchange information regarding the characterization and application of the commercial imagery used by the government. This workshop is focused on the synergy of high, medium and low resolution imagery and remote sensing technologies used by the Government. This workshop is sponsored by the Joint Agency Commercial Imagery Evaluation (JACIE) team, a collaborative group of representatives from the National Aeronautics and Space Administration (NASA), the National Oceanic and Atmospheric Administration (NOAA) the United States Department of Agriculture (USDA) and the United States Geological Survey (USGS).

Tell Me About ASPRS 2014

The conference theme: Power in Our Pockets, refers to the technological power of pocket sized devices in our world today. The conference will focus on the various tools, applications, software and overall abilities of technology in the geospatial industry today.

There are LOTS of changes happening for ASPRS conferences! Here are just a few you will see in 2014:

  •     JACIE Workshop co-location
  •     Unmanned Aerial Systems Showcase
  •     Recruitment Way Table Tops
  •     Increased Exhibitor/Attendee Face-time
  •     New session tracks for practical applications
  •     Redesigned conference programs
  •     Presenter abstracts available online

Who Attends?

More than 1,000 imaging and geospatial information professionals gather from across the nation and from around the globe for ASPRS Annual Conferences. And this year, we are expecting a record attendance with the co-location of the JACIE Workshop.

Attendees are mid- and upper-level imaging and geospatial managers from corporations, government agencies, consultants, educators, reseachers, students and field surveyors.


Louisville, Kentucky
Big City Service with Southern Hospitality!

Nestled on the banks of the Ohio River, Louisville, Kentucky has loads of small-town southern hospitality, a cosmopolitan riverfront district linked to the city’s park system, a diverse arts scene, downtown’s Museum Row on Main, and a nationally recognized foodie mecca.

Louisville, no matter how you pronounce's got something for everyone!

ASPRS and the JACIE Workshop will be holding meetings at the wonderful Galt House Hotel while in Louisville! Click here for more information about hotel accommodations.


T3-Mar 25 9:15

Small Business Size Standard Increase Panel Discussion

Brian Murphy

A Post-Mortem on the 2012 SBA Small Business Size Standard change from $4.5M to $14M.  I would like to organize a Panel Discussion regarding a post-mortem on the SBA's increase of the Small Business Size Standard for NAICS Code 541370 - Surveying and Mapping (except Geophysical) Services  One - Two Small Business companies who were/are under the $4.5M threshold. One - Two Small Business Comapnies who became a small business after the size standard change  One to two large businesses above the $14M size.

T9-Mar 27 11:00

Extraction of planar and pole-like road features from terrestrial mobile laser scanning data

Zahra Lari, Department of Geomatics Engineering, University of Calgary

Ayman Habib

Road network infrastructures are considered as one of the key elements for the economic growth and social cohesion for a given society. Therefore, periodic monitoring of their condition is essential for the inventory, performance evaluation, and visibility analysis. To date, different monitoring systems have been developed for the road infrastructures surveillance. However, the majority of them are manual, error-prone, and temporally-sparse. In recent years, terrestrial mobile laser scanning (TMLS) systems have been acknowledged as an efficient and reliable alternative technology for the collection of high density 3D point clouds along the road corridors.  The acquired point clouds should be processed for the extraction of useful road asset information. This paper introduces a novel automatic approach for the identification, parameterization, segmentation, and extraction of planar (e. g., road surface, road signs, curbs) and pole-like (e.g., light poles, traffic poles) features from TMLS data. The proposed method takes the internal characteristics of the collected data - i.e., local point density variations and noise level - into account. In the first step of this approach, the points which belong to planar and pole-like features are detected and represented based on the principal component analysis of their local neighborhood. For the detected planar features, the segmentation attributes are then estimated using an adaptive cylinder neighborhood defined based on their appropriate representation model. Afterwards, an adaptive clustering technique is implemented to discriminate individual planar features in the parameter domain. For the pole-like features, the representing direction, position, and radius parameters are utilized as the segmentation attributes. A parameter-domain clustering method is finally utilized to isolate the points which belong to individual pole-like features in direction, position, and radius attribute spaces, successively. Experimental results from simulated and real data will demonstrate the feasibility of the proposed approach for the extraction of road features from TMLS data.

Region Growing Approach for the Extraction of Cylindrical/Linear Features from Laser Scanning Data

Kaleel Al-Durgham, University of Calgary

Ayman Habib

Laser scanning systems have been repeatedly utilized for the direct acquisition of three-dimensional spatial information since it provides high density and accurate 3D point cloud over physical surfaces. The raw point cloud does not provide information about the type of the scanned surface and therefore further segmentation and identification steps should be applied over the scans. Cylindrical and linear features are examples of the features that can be automatically extracted from laser scans. The proposed work will present a reliable framework for the automatic extraction of cylindrical/linear features from laser scanning data. Different factors such as the noise level and the local point density will be considered for the utilized set of points within a local cylindrical neighborhood for the point under consideration. For a certain neighborhood, the local surface properties can be determined through Principal Component Analysis, and thereafter different geometric features within the scan can be identified. Points that are identified as a part of cylindrical/linear features are used as seed points to start a region growing process. An appropriate modeling of cylindrical/linear features is utilized where a minimum parameterization of cylindrical/linear features is considered. Experimental results from a power plant dataset illustrate that the proposed methodology can be utilized for the automatic extraction of cylindrical/linear features from laser scanning data where pole-like objects and power lines were successfully identified and extracted.

Mapping of levee lines using LiDAR Data and multispectral orthoimages

Yunjae Choung, The Ohio State University

Ron Li

The mapping of levee lines is important for detailed description of levee shapes and for detection of topographic changes to the levee surface. This research proposes a method for the mapping of levee lines using aerial LiDAR data and the multispectral orthoimages. First, multiple algorithms have been developed including algorithms for breakline detection, morphological filtering, modified convex hull and polynomial approximation algorithm. These are used to extract levee lines from LiDAR data (thereafter called the LiDAR-based lines). The introduced method can identify various types of levee lines including those that are located at the step, ramp and/or curved edges of the levee top and toe. After the LiDAR-based lines have been extracted, the levee edges are extracted from the image sources using Gaussian filtering with two parameters. These parameters can include the original intensity value of the input image (Parameter 1) and the distance values (Parameter 2) measured from the pixel of the input image to the LiDAR-based lines. From among the set of edges extracted from the image sources, the appropriate edges can be selected and identified as the levee edges. These selected levee edges are primarily used as the levee line segments. In areas where the levee edges could not be extracted, the LiDAR-based lines can be used as the levee line segments. Finally, the complete levee lines are constructed by connecting these segments, whether the levee edges extracted from the image source or the LiDAR-based lines.

Sustainable Geotechnical Asset Management along the Transportation Infrastructure Environment Using Remote Sensing

Rudiger Escobar-Wolf, Michigan Tech University

Thomas Oommen, Colin Brooks, Pasi Lautala, and Stanley Vitton

With the aim to develop a novel approach for the management of geotechnical assets (e.g. retaining walls, unstable slopes, rockfall sites, cut slopes, and embankments) a new approach using commercially available remote sensing is proposed, to provide a cost-effective solution for the transportation agencies to measure the state of the geotechnical assets using remote sensing, relate that data to obtain information on the condition of the asset, and further utilize this information for strategic investment to achieve life-cycle performance goals of the asset. The current management practices for geotechnical assets along the transportation environment have been mostly restoring the asset after the failure, instead of identifying and remediating hazardous conditions before their occurrence. The reason for lacking a proactive geotechnical asset management system is that the geotechnical assets are extensive and assessing their condition using traditional site inspections by trained engineers is laborious and costly.  Recent advancements in commercial remote sensing (InSAR, LiDAR, and optical) provide opportunities to obtain precise measurements of displacements and these displacement measurements could provide a valuable alternative to the traditional laborious site inspections to determine the condition of the geotechnical asset. Measurements and asset conditions can be integrated into a decision support system for assisting asset managers to make decisions on short and long-term investments. The project team will solicit input from state Department of Transportations (DOTs) and other transportation agencies to define the requirements for the remote sensing based geotechnical asset management system.

Geospatial Power Applications to Developmental Projects Monitoring and Evaluation to Eliminate Corruption in Nigeria

Matthew Adepoju, National Space Research and Development Agency

Halilu Shaba, Mohammed Seidu, and Alade Taslim

A large number of developmental and infrastructural projects have been budgeted and awarded for many years raging from roads construction, schools, and hospitals to irrigation dam and many more. Many of the contractors have colluded with government officials to get full payment or substantially mobilised while little or nothing has been achieved in many of such laudable projects. This has left the country with near total collapsed of the basic infrastructure leading to poor standard of living, unemployment, and many times lost of lives and properties. This paper proposed a paradigm shift of government policy on monitoring and evaluation to employ geospatial power to monitor, evaluate and manage Federal Government projects in Nigeria to ensure the realisation of the Government Transformation Agenda and Vision 20: 2020. The attainment of this has been totally shackled due to the monumental level of corruption in all sectors made possible for lack of appropriate technological utilisation. Some selected projects revealed overwhelming and monumental corruption in the award, implementation, monitoring and evaluation of these projects. The application of satellite remote sensing provided a synoptic view as well as temporal profiling of the stages of work, location, and progress that were made over a period of 10 years compared with budget release of fund. Geospatial power provides unique opportunity to tackle these challenges because large areas are covered with satellite data and provides high temporal resolution (time series) as well as spatial resolution suitable for this application. In addition, GIS made evaluation less dependent on human interference thereby made the systems robust with functionalities and operational capability for mapping, monitoring and profiling of changes taking place at the projects sites that are often remotely located. The result shows the urgent need for Government Budget Office to make payment according to the level of work done. It is recommended that an immediate development and deployment of location-based integrative infrastructural monitoring and management systems be adopted by Federal Government through Central Bank on payment for contracts, and policy review to ensure effective and timely implementation of projects in Nigeria.

GUI-based Image ANnotation Tool (GIANT)

Paul Pope, Los Alamos National Laboratory

Doug Ranken and Meilin Yan

The GUI-based Image ANnotation Tool (GIANT) is software for rapidly annotating an image.  GIANT provides an easy-to-use graphical user interface (GUI) that allows a user to load an image and immediately begin to create point-based labels, which are displayed as an overlay on the image.  Both grayscale and color Tagged Image File Format (TIFF) images can be displayed and annotated. The labels (feature names) are derived from a controlled vocabulary, which can be customized by the user to fit any land cover/land use theme or knowledge domain.  Annotation changes (additions or deletions) are continuously and instantly recorded to a GIANT Annotation File (GAF), so that saving or reloading the annotation information occurs transparently to the user.  GIANT provides rapid image zoom, scroll, pan, and rotation capabilities.  It also continuously updates an image-to-display transformation, whose inverse is used to convert user-selected annotation locations from display coordinates to image coordinates.  The feature names and row/column image coordinates of the label locations are recorded to the GAF file.  GIANT provides a tool so that this information can be converted to a Comma Separated Value (CSV) file.  The annotations can then be used to support other analyses, such as cartography, imagery analyst photo-interpretation comparisons, and the training/testing of semantic feature extraction algorithms.  GIANT was written using the Interactive Data Language (IDL), a computer programming language developed and sold by ExelisVIS.  GIANT has cross-platform capability, and can be run using IDL’s license-free “Virtual Machine,” or from the IDL command-line, or the ENvironment for Visualizing Images (ENVI).

Community Driven Asset Maps

Alex Bostic, URS

The Tampa Bay Partnership is responsible for driving business to the Tampa Bay Florida region.  One of the ways they do this is by highlighting the capabilities of various sectors such as Applied Medicine and Human Performance and High-Tech Electronics and Instruments.  For the last few years, the Tampa Bay Partnership had been contemplating the use of maps to help visualize various assets like Workforce Development, or Medical Services from a variety of sectors.  The Tampa Bay Partnership met a few challenges along the way, but with the emergence of ArcGIS Online their vision has come to fruition.  Come learn how three of the Partnership’s workflows of viewing sector assets, crowd sourcing new assets, and verifying the crowd sourced assets were implemented with ArcGIS Online.  We’ll discuss the process that was taken in order to make their vision of “Asset Maps” a reality.

Using Cloud technology to deliver geospatial data and tools to the enterprise

Stephen Ellis, GISP, CMS, AeroMetric

The presentation will focus on geospatial cloud solutions for multi-modal imagery management and distribution. The cloud-based solutions are supported by Esri's ArcGIS and can be quickly put in operation. In addition to dedicated portals, the system supports standards-based geospatial web services that enable users to manage and access geospatial information on their own terms with their own tools.

Modeling Tallgrass Prairie Above-Ground Biomass in the Central Great Plains Using Hyperspatial Multispectral Imagery

Chuyuan Wang, Department of Geography, Kansas State University

Kevin P. Price, Deon van der Merwe, Huan Wang, and Nan An

The objective of this research was to examine the relationship between tallgrass prairie above-ground biomass (AGB) measurements and hyperspatial multispectral imagery collected by small unmanned aircraft systems (sUAS) using NDVI as the predictor. AGB measurements were sampled across a biomass gradient using 12 sampling frames in the Tallgrass Prairie National Preserve in Chase County, KS. sUAS imagery was collected over sampling frames at three low altitudes by both a modified Canon S100 NDVI camera and a modified Canon T4i NDVI camera equipped under an unmanned DJI S800 hexacopter. Aerial photos for the whole study area were collected by a modified Canon 6D NDVI camera at about 2,000 m above the ground flying in a piloted jet aircraft. Tallgrass AGB measurements were then analyzed against NDVI data derived from imagery of different altitudes using regression analysis. Results show that there is a statistically significant exponential relationship between NDVI values and tallgrass AGB measurements. The model built using NDVI data derived from the lowest altitude imagery can be used to explain over 95% of the total amount of variation in biomass data, but R-squared value decreases as flying altitude increases. It means that spatial resolution does have a significant effect on the model’s predictability. There is also a very strong correlation (R2>0.90) between each two sets of NDVI data collected at different altitudes, which demonstrates that there is a great potential to upscale the biomass information obtained from sUAS imagery to satellite imagery of much coarser spatial resolution to make tallgrass prairie biomass predictions over much larger geographic area.

Automatic detection of zebra crossings from mobile mapping images

Yongjun Zhang, School of Remote Sensing and Information Engineering,Wuhan University

Lu Hongshu, Xiong Xiaodong, Wang Bo, and Wang Qing,

Zebra crossing play an important role in the regulation of traffics and the protection of pedestrians. The rapid expansion of cities leads to dramatic increasing of vehicles, which makes the zebra crossings fade insensibly. Traffic accidents would occur if the faded zebra crossings were not repainted in time. However, there is no standard to judge the appropriate time for repainting. This paper aims at the automatic detection of fading status of zebra crossings based on optical images obtained by Mobile Mapping System (MMS).   In this paper we introduce an artificial intelligence method to detect zebra crossings and vehicles automatic. In this method we extracted Haar and LBP features from positive and negative samples to train the classifier. Then we estimated the probability of fading through the detection results. The combination of the two features is used in order to guarantee robustness. Furthermore, we can establish database to record the positions of zebra crossings,the number of zebra-crossings and some comprehensive values about the probability of zebra-crossing fading in one district. This database can provide a reference for road management by indicating when and where the zebra crossings need to be repainted. This detection method provides a comprehensive and objective standard to judge the appropriate time to repaint. And the input data can coverage a wide district with little cost and be updated easily during a period.

Automated edge-based extraction of power lines from imagery of electrical substations

Hossein Armeshi, University of Calgary

Ayman F. Habib

Electrical substation is one of the most important components of electrical systems in the world. Unfortunately, some electrical equipment in electrical substations is propitious place for some animals to perch, roost, and hunt. Accordingly, animal electrocution and power outage may easily happen in electrical substations. Researchers have proposed different solutions to protect animals from electrocution in electrical substations. Covering the electrical equipment is one possible solution which requires field work to measure their dimensions. To avoid dangerous field work within high voltage environments, remote sensing-based measurement seems to be an appropriate alternative. Among the equipment of electrical substations, power lines are of interest to this paper. In this paper, a new feature-based algorithm is introduced for the automated extraction of power lines from imagery of electrical substations. These features include those derived from detected edges. Moreover, Principal Component Analysis is utilized to prepare edges for linear feature extraction. Furthermore, geometrical properties of the power lines are employed to extract and group segments of the power lines. Finally, linear features which have been wrongly selected as power line are eliminated by considering intrinsic characteristics of electrical equipment.  The proposed algorithm is applied to a wide range of images taken from different scenes of electrical substations. The preliminary results demonstrate high efficiency in extracting the power lines. However, some work is still needed to increase the reliability of automated extraction of the power lines. For example, the extracted power lines from the proposed algorithm seems to provide proper information as training data for some supervised classification methods to extract all power lines.

An Application of Object-Based Land-Cover Mapping for Tippecanoe County, Indiana

Li Xiaoxiao, Arizona State University

Detailed land-cover mapping is essential for a range of research issues addressed by the sustainability and land system sciences and planning. Large-scale land cover maps, such as National Land-Cover Dataset (NLCD) and National Agricultural Statistics Service (NASS), have coarser resolution than some currently available aerial photography and satellite imagery, and traditional pixel-based classification methods alone are not enough for land-cover mapping in the complicated, spectrally heterogeneous urban areas. In this study an object-based approach was utilized for identifying land-cover types from 1-meter resolution aerial orthophotography. The aerial photography was pre-processed by using principal component transformation to reduce its spectral dimensionality. Vegetation and non-vegetation were separated via masks that created with the Normalized Difference Vegetation Index (NDVI). A single input dataset was employed with simple algorithms to minimize computational time and to maintain high accuracy in land cover mapping. Different segmentation algorithms with lower calculation intensity were combined to generate image objects that fulfill feature selection needs.  Expert knowledge was employed in image object identification process. Characteristics of image objects from contextual and geometrical aspects were concerned in the decision rule set to reduce the spectral limitation of the four-band aerial photography. The accuracy of land-cover mapping by using the object-oriented approach was higher than pixel based classification. 

You are here: Home Louisville 2014