Dr. Walter Scott, DigitalGlobe
Dr. Walter Scott is the founder of DigitalGlobe and currently serves as Executive Vice President and Chief Technical Officer. He is responsible for DigitalGlobe’s Platform and Services Business Units, as well as space system acquisition. Scott holds a Bachelor of Arts in Applied Mathematics, magna cum laude, from Harvard College and a Doctorate and Master of Science in Computer Science from the University of California, Berkeley.
VADM Robert B. Murrett (Ret)
Robert B. Murrett is a professor on the faculty of the Maxwell School of Citizenship and Public Affairs at Syracuse University. Previously, Murrett was a career intelligence officer in the U.S. Navy, serving in assignments throughout the Pacific, Europe, and the Middle East through his thirty-four years of duty, retiring in the grade of Vice Admiral. For the last ten years, he served as Vice Director for Intelligence, U.S. Joint Chiefs of Staff, Director of Naval Intelligence, and Director of the National Geospatial-Intelligence Agency (NGA).
Lawrie Jordan, Esri
Lawrie Jordan is the Director of Imagery for Esri, as well as Special Assistant to Jack Dangermond, President of Esri. Mr. Jordan has more than 30 years of experience as a leader in the field of image processing and remote sensing, including a long standing strategic partnership with Esri. His background education is in Landscape Architecture, with degrees from the University of Georgia and Harvard University.
Attendees will glean knowledge and ideas from other successful real-world analytical applications, all while sitting in the beautiful surroundings of the Brown Palace Hotel & Spa.
Attendance for the EAS is being limited to 200 people in order to provide high-quality interaction and participation.
The theme of the 2017 ENVI Analytics Symposium (EAS) is “Analytical Solutions in a Data-Rich World.” With the staggering volume of commercial and open-source data that is becoming available, organizations will need to quickly transition to new business and service models to be successful. Agility, technology, and innovation will separate the winners from the losers. Register to attend the 2017 EAS to keep your finger on the pulse of this fast-growing and dynamic market of commercial geospatial Big Data analytics.
The ENVI Analytics Symposium is being held at the Brown Palace Hotel and Spa in Denver, CO. The Brown Palace is a Forbes 4 Star hotel and is a legend among Downtown Denver hotels. Brown Palace guests enjoy access to timeless luxury with a unique sense of place, original experiences and world-class service and amenities. There's simply no better way to experience the Mile High City.
Registration is limited, so don't wait 'til it's too late.
We have multiple pricing options to best fit your needs. Click below to get the details.
Ursa Space Systems
Julie Baker is Co-founder and VP of Operations at Ursa Space Systems. Julie has 30 years’ experience in the software industry, including 15 years in various software engineering roles and 15 years in technical management. Prior to co-founding Ursa Space Systems, she was VP of Cyber Technology at Architecture Technology Corporation where she provided technical leadership and management of the company's research and development in the areas of cyber security, information management and reliable computing. Julie holds a Master of Science in Computer Science from Stanford University and a Bachelor of Music from the University of Texas at Austin.
John Corbett, Ph.D., agricultural climatologist and aWhere, Inc. Co-Founder and Chief Science Officer, focused his career on applied agricultural meteorology and a scalable, enterprise, agricultural intelligence platform. aWhere maintains a global resource of high-resolution, current, daily observed and forecast agronomic weather data for localized ‘smart content’ planet-wide. With 25+ years working in agriculture, John leads a team of software experts, agricultural scientists, and business leaders leveraging technology for actionable, timely, location specific information meeting needs across the agricultural value chain. Prior to aWhere, John worked with Syngenta-Switzerland, Texas A&M- USA; with ICRAF – Kenya; and with CIMMYT – Mexico.
Owen Cox is a Programming Consultant within the Custom Solutions group at Harris Geospatial UK. Working within the UK team since 2012, Owen works with Harris’ COTS products and third party technologies to realise bespoke solutions to user problems, principally focused on Enterprise technologies. Previously he was a Remote Sensing Scientist at the UK Met Office, working on novel uses of meteorological Lidar and Radar.
Brian Curtiss is one of the founders of ASD, Inc. and serves as Chief Technical Officer. He has over 30 years’ experience in the field of spectroscopy and optical sensing and holds a B.A. in Earth and Planetary Sciences from Washington University in St. Louis and a M.S. and Ph.D. in Geochemistry from the University of Washington. As a post-doctoral research fellow at CalTech, he assisted in the development of the field portable spectroscopic instrumentation. Prior to joining ASD, he held a research faculty position at the University of Colorado in Boulder. After co-foundering ASD Inc. in 1990, Dr. Curtiss served as principal investigator on seven NASA, NSF and DOE funded research projects. In his current position as ASD’s Chief Technical Officer, Dr. Curtiss applies his experience in the fields of spectroscopy and optical imaging to a diverse range of analytical problems in the natural resource industries.
John has over 40 years of experience establishing enterprise architectures with video broadcast solutions, networking, asset management, video production, and geospatial enterprise systems. He was the chief architect for Harris platform solutions which have been deployed across the DOD and IC. He is currently the Co-Chair for the USGIF ABI Working Group.
John has extensive experience in cloud computing having been instrumental in the architecture of Harris H10 Multi-Domain Cloud Enterprise Architecture which leverages OGC Rest and SOAP services. He is also leading Harris Next Generation of Geospatial Enterprise solutions which will enable multi-source discovery, provide a commercial analytics marketplace, and model driven analysis platform.
Previous to joining Harris Geospatial business John was the Chief Strategy and Technology officer for Harris Commercial Broadcast Division. Throughout his career he led efforts to build the world’s first Digital AM and FM Radio systems, Digital Television, IP Video networks and implemented Harris Cross Platform Advertising Campaign Management Platform. He pioneered Harris FAMEtm Geospatial Digital Asset Management System, among the first digital geospatial video content management data models. He is also considered a leading expert in CDN distribution architectures, video encoding, video analytics, video networks and modern geospatial enterprises.
John is well known as a speaker on emerging technologies. He holds a BSEE in RF Engineering and BSEE in Broadcast Engineering, is a Graduate of Darden UV Executive Business Innovation Program, Graduate of Kellogg Executive Management Program.
Lawrie Jordan is Director of Imagery and Remote Sensing for Esri, as well as Special Assistant to Esri founder and President, Jack Dangermond. Mr. Jordan has over 35 years of experience as a leader in the field of image processing and remote sensing, and played a key role in evolving a long standing strategic partnership with Esri. His background education is in Landscape Architecture, with degrees from The University of Georgia and Harvard University.
Lawrie is a member of the European Academy of Sciences and Arts, as well as the recipient of the Geospatial World Leadership Lifetime Achievement Award for his decades of contribution in the field of Image Processing and Earth Observation. He is also grateful to be the recipient of the U.S Government’s medal for Outstanding Support and Patriotism.
Mr. Larson has over 20 years of space sector experience with the Canadian Space Agency, MacDonald, Dettwiler and Associates (MDA), and most recently as President and COO of Urthecast, of which he was a Co-Founder. He also has extensive experience in space-related strategy formulation, business
development, government relations, corporate development, and operations.
Rebecca Lasica is the Sr. Solutions Engineer Manager at Harris Corporation and has been immersed in software and remote sensing science for more than ten years. She is a University of MN alumni with industry expertise focused on satellite, airborne, and UAS image analytics.
Daniela Moody Ph.D.
Daniela’s work at Descartes Labs focuses on developing improved feature extraction algorithms for multispectral satellite imagery that combine sensor fusion, adaptive signal processing, and machine learning techniques. She was at Los Alamos National Laboratory for 9 years prior to joining Descartes Labs, working on remote sensing and machine learning applications in various research areas, including space systems, astronomy, and nuclear non-proliferation. She received her M.S and Ph.D. in Electrical Engineering from the University of Maryland, College Park in 2012.
Robert B. Murrett
Robert B. Murrett is a professor on the faculty of the Maxwell School of Citizenship and Public Affairs at Syracuse University, and serves as the Deputy Director of the Institute for National Security and Counterterrorism (INSCT) at the University. He is also a member of the Board for the Institute for Veterans and Military Families, and is responsible for a series of ongoing research projects between the University and the Syracuse Veterans Administration Medical Center. He is on the adjunct staff of the RAND Corporation, the Institute for Defense Analyses, and chairs the MITRE Intelligence Advisory Board.
Previously, Murrett was a career intelligence officer in the U.S. Navy, serving in assignments throughout the Pacific, Europe, and the Middle East through his thirty-four years of duty, retiring in the grade of Vice Admiral. His duty stations included service as Operational Intelligence Officer for the U.S. Pacific Fleet, Assistant Naval Attaché at the U.S. Embassy in Oslo, Norway, and Director for Intelligence, U.S. Joint Forces Command. For the last ten years, he served as Vice Director for Intelligence, U.S. Joint Chiefs of Staff, Director of Naval Intelligence, and Director of the National Geospatial-Intelligence Agency (NGA). He holds an undergraduate degree from the University of Buffalo, and a masters degree from the Walsh School of Foreign Service at Georgetown University.
Nazlin Kanji is a Product Line Director at AeroVironment. She has more than 20 years of experience developing and deploying big data systems for the DoD and commercial industry, and is now leading AeroVironment’s commercial data programs focusing on agriculture, utilities, transportation and energy. She was the program manager on the first FAA-approved commercial unmanned aircraft operations over land and water in the North Slope region and had the pleasure of spending her summer in the Arctic supporting the operations. She has a BS in Computer Science from California State University, Northridge and earned an MBA at California Lutheran University.
Paolo has been enjoying the last 25 years investigating the many aspects of extracting useful and reliable information from Synthetic Aperture RADAR imagery, developing algorithms and exploring different applications. After realizing in his years in the academia at Politecnico di Milano and University of Zurich that these goals are not just dreams, he co-founded sarmap to bring this experience into operation, leading the development of the SARscape software package.
Pedro Rodriguez Ph.D.
Dr. Pedro A. Rodriguez is the Senior Technical Leader of multiple Deep Learning projects at the JHU Applied Physics Laboratory. His work includes developing Deep Learning algorithms for automatic target recognition on a variety of sensor modalities such as: Electro-Optical, Infrared, Synthetic Aperture Radar and Full Motion Video. Dr. Rodriguez holds a M.S. in Applied Biomedical Engineering from Johns Hopkins University and a Ph.D in Electrical Engineering from the University of Maryland, Baltimore County. He has more than 12 years of experience developing novel image detection, tracking, classification and fusion algorithms for a variety of Information, Surveillance and Reconnaissance projects.
Mark E. Romano
In his capacity as the Sr. Product Manager for the Harris Geospatial Solutions Division, Mr. Romano is responsible for commercialization of Geiger mode LiDAR and other space, air, land, and sea remote sensing capabilities and services. Mark has an extensive background with 30+ years’ experience working with defense, Fed/civil, and commercial communities as a recognized industry expert, leading development of innovative and disruptive remote sensing technologies. He has authored and co-authored numerous papers, journals, text books, and specifications for government and industry with active participation in panels and committees as a subject matter expert.
Dr. Walter S. Scott
As Executive Vice President and Chief Technical Officer of DigitalGlobe, Dr. Scott oversees the development of space systems, R&D, and DigitalGlobe’s Platform and Services Business Units.
Dr. Scott founded DigitalGlobe in 1992 as WorldView Imaging Corporation, which was the first company to receive a high resolution commercial remote sensing license from the U.S. Government (in 1993), under the 1992 Land Remote Sensing Policy Act. WorldView became EarthWatch Incorporated in 1995. Dr. Scott managed the development of all of the company’s commercial remote sensing satellites. He secured the first-ever export license for launch of U.S.-manufactured imaging spacecraft on Russian launch vehicles (Start-1 and Cosmos). The company became DigitalGlobe in 2001, and with the launch of the QuickBird-2 satellite that year, offered the world’s highest resolution commercial satellite imagery. Today, DigitalGlobe operates a 5-satellite imaging constellation with the best revisit and greatest capacity in the industry.
From 1986 through 1992, Dr. Scott was with Lawrence Livermore National Laboratory (LLNL). He began as Project Leader for Computer Aided Design Tools for the Laser Pantography Program, developing tools to aid in the design of wafer scale integrated circuits manufactured. In 1987, he joined a small team developing a concept for a highly distributed constellation of space based interceptors for the Strategic Defense Initiative, known as “Brilliant Pebbles.” In late 1987, Dr. Scott became Program Leader for this effort, responsible for creating a series of hardware prototypes and conducting flight experiments. During 1989, Dr. Scott led the program successfully through over 20 reviews of technical feasibility, system performance, military operability, and estimated cost, resulting in the adoption of Brilliant Pebbles for SDIO’s space segment in 1990. In late 1991, Dr. Scott was Assistant Associate Director of the Physics Department and was responsible for development of new space-related programs and identification of promising technologies.
Prior to joining LLNL, Dr. Scott founded and served as president of Scott Consulting, a UNIX systems and applications consulting firm. He developed Unix networking subsystems, and a pioneering email system that used public key encryption for message protection.
Dr. Scott holds a Bachelor of Arts in Applied Mathematics, magna cum laude, from Harvard College and a Doctorate and Master of Science in Computer Science from the University of California, Berkeley. He was a visiting student for a year at Edinburgh University in Scotland.
Dr. Scott was named Entrepreneur of the Year by Ernst & Young in 2004 for the Rocky Mountain Region in the Emerging Technology category.
He previously served on the National Research Council’s Committee on Earth Science and Applications from Space (CESAS) and is currently member of the board of directors of the Open Geospatial Consortium (OGC).
Alex Shih is Director of Product & Ecosystem at Planet. Planet designs and manufactures miniature satellites with the mission to image the entire world every day, and make global change visible, accessible, and actionable.
He brings experience from the SaaS, cloud, and mobile space, having previously led mobile partner products at Twitter and helped launch the enterprise partner program at Google for Google Apps.
Alex has degrees from MIT and Cornell University.
MDA Information Systems
Dr. Douglas S. Way is Chief Scientist, MDA Information Systems Geospatial Solutions and Professor Emeritus at The Ohio State University. Prior to joining MDA Federal in 2004, Dr. Way was Professor and Department Chair at the Knowlton School of Architecture, Ohio State University and was previously Professor at the Harvard Graduate School of Design, Harvard University. His over 45 years of research and professional practice has focused upon remote sensing and geographic information systems spatial modeling applied to land use change dynamics and a wide variety of environmental and strategic issues related to U.S. national security. Dr. Way earned his B.S. at the University of Wisconsin, M.A at Harvard University and M.A. and PhD. in geography-geomorphology at Clark University and is author of the text Terrain Analysis.
University of California, Santa Barbara
Erin Wetherley is a doctoral candidate in the Geography Department at the University of California, Santa Barbara. Her work with Drs. Dar Roberts and Joe McFadden focuses on characterizing urban climate variability using imaging spectrometry, thermal imagery, climate modeling, and sub-pixel analyses. Additional research interests include mapping post-fire landscape recovery and measuring the spectral characteristics of vegetation drought response. Prior to her doctoral work, Erin earned a bachelors degree in Environmental Studies from Brown University, and worked for several years as a GIS and database manager at a Washington, D.C. non-profit organization.
Click here to view workshop titles and abstracts
Improving Outcomes with UAV and Airborne Data
Over the years, Harris Geospatial has worked with many customers that use data collected on UAV and airborne platforms. This session will dive into lessons we’ve learned with regard to:
Applied Machine Learning
Recently, “Machine Learning” has been the focus of many conversations in the remote sensing community. However, Machine Learning has had varying degrees of quality, testing, and ultimately results. Harris has developed a Machine Learning solution that is mature, deployable, and capable of scaling for large projects across modalities including multispectral, hyperspectral, and LiDAR. In this workshop, high-level principals of Machine Learning will be discussed and Machine Learning will be shown to be a leap over traditional pixel-based analytics for some features and data. Several use cases will be presented and there will be discussions about how you can take advantage of this technology.
Deploying to Partner Platforms
Harris is proud to partner with many of the leading data and platform providers who are important contributors to our industry as a whole. These partnerships enable us to deploy analytics in close proximity to the data, and expose them to user bases across several different industries. In this workshop, Harris will present several deployments of enterprise geospatial analytics on various partner platforms. This will be an engaging session where attendees will gain technical insight into how these analytics are deployed with hands-on opportunities to interact with the platforms.
Vegetation Analysis: Why it’s Hard and What You Can Do
Remote sensing of vegetation has been around for as long as remote sensing, but many are still confounded on how to observe vegetation condition, species, and change. This workshop will give a background on vegetation remote sensing, how to improve outcomes for observation, and modalities for best observing vegetation phenomenology. The new ENVI Crop Science product will be presented and there will be time for Q&A with vegetation experts from Harris.
Data Modalities, Sources, and Achieving the Best Results
Today, there is more data than ever before, and it is critical to derive meaningful and actionable information from this data. In this workshop, Harris will discuss recent data modalities that are now available such as small sats, weather sensors, and emerging commercial platforms. Some of these data are already available in cloud storage and delivery environments. We’ll show you how to ingest, work very large datasets efficiently, and describe potential best applications for the different sensors. This workshop will also discuss potential applications and delivery of machine learning.
Real-time Problem Solving for the Enterprise and Desktop
Bring your problem solving skills and geospatial expertise to this workshop for some hands-on ENVI analytics fun. Come ready to get creative and solve a real geospatial remote sensing problem, complete with access to data and ENVI tasks and workflows. Attendees should be somewhat familiar with remote sensing phenomenology and/or optical image analytics. Advanced users will have access to all the tools needed to solve these problems.The sky’s the limit.
Click on a title and speaker to get a more in-depth abstract of each presentation.
SESSION 1 : THE FUTURE OF EARTH OBSERVATION
SESSION 2 : GEOPROCESSING APPLICATIONS AND PLATFORMS
This discussion and demonstration will focus on mapping and analyzing unstructured social media textual data using GIS technologies. Social media mapping is a very hot topic. Many social media feeds and connections provide access to GPS-geolocated social media, which can be mapped in a variety of ways based on geolocation information. However, most unstructured social media content is not geolocated. ClearTerra LocateXT technology provides the ability to find and map location-based information within the unstructured data of social media even without GPS-based geolocation data. Along with mapping, the ability to tag locations based on keywords and content discovered in proximity of the location (for example, activities occurring or entities present) will be shown. These workflows will be demonstrated using web-mapping GIS services hosted in the cloud. Examples will be shown that would be particularly effective for intelligence and law enforcement scenarios.
Increased weather variability driven by a warming atmosphere impacts food production. Human security in some areas is already showing resource constraint pressure and this will only grow more intense. Monitoring agricultural production provides detailed localized information suitable to inform crises intervention efforts and optimize and target appropriate interventions.
Activity Based Intelligence (ABI) enables analysts and decision makers, using multi-source intelligence to develop situational awareness based on human activity. To effectively contribute to ABI, imagery intelligence (IMINT) needs to derive features from the imagery and store them in spatio-temporal databases enabling ease of access, pattern identification and fusion with intelligence from other sources.
This presentation will describe an enterprise processing system to automatically extract ABI from multi-source image data. Developed by Harris for UK defense, it utilizes Harris’ Geospatial Services Framework to provide a framework where scientists can deploy algorithms which are then intelligently applied to incoming image data. Detected features are then automatically stored in a geospatial database, enabling subsequent analyst exploitation in dashboards or via visual inspection.
We will demonstrate how to use field reflectance spectra, collected for materials such as vegetation, as input into calibration models that can be applied to map quantitative information in hyperspectral images. For example, quantitative results from using this technique for vegetation can provide information about leaf chemical properties like canopy nitrogen and lignin content. The resulting calibrations for vegetation properties produced using this method can then be mapped to hyperspectral images. Indices like NDVI are useful, but they lack the ability to give detailed information on physiological processes, so this technique takes image analysis a step further. This type of analysis can be very important in many different applications in areas like precision agriculture, forestry, environmental monitoring, mining and defense and intelligence.
SESSION 3 : INDUSTRY SOLUTIONS AND GEOSPATIAL ANALYTICS
Operating with the highest levels of efficiency, reliability and safety is a top priority for utilities. The growth of distributed generation and diversification of power sources bring operational system challenges and an aging infrastructure and workforce is driving the need for asset renewal prioritization and knowledge capture. Using the combination of condition-based maintenance and predictive maintenance, utilities can effectively overcome these challenges and remain relevant in the changing energy marketplace. Incorporating the use of drones in an intelligent, condition-based, asset management program will provide utilities with the information needed to operate more efficiently, effectively, and safely, consequently allowing them to overcome some of these disruptive obstacles. Utilizing data collected from drones will lead to reduced truck roll outs and ensure the right assets and skill sets are deployed. Additionally predictive maintenance utilizing drones ensures assets remain in working order, reducing failures in the grid, especially where a potential asset failure could result in significant damage. The addition of drone data in the utilities industry has the potential to revolutionize the industry.
The recent computing performance revolution has driven improvements in sensor, communication, and storage technology. Multi-decadal remote sensing datasets at the petabyte scale are now available in commercial clouds, with new satellite constellations generating petabytes/year of daily high-resolution global coverage imagery. Cloud computing and storage, combined with recent advances in machine learning, are enabling understanding of the world at a scale and at a level of detail never before feasible. We will present results from an ongoing effort to develop satellite imagery analysis tools that aggregate temporal, spatial, and spectral information that scale with the high-rate and dimensionality of imagery being collected. We will focus on the problem of monitoring food crop productivity across the Middle East and North Africa, and show how an analysis-ready multi-sensor data platform enables quick prototyping of various satellite imagery analysis algorithms.
Today, there is more data than ever before, and it is critical to derive meaningful and actionable information from these data. In this workshop, Harris and some participating partners will discuss recent data modalities now available such as data from small sats, weather sensors, and commercial platforms. Some of these data are already available in cloud storage and delivery environments. We’ll show you how to access, ingest and work with very large datasets efficiently and also describe the potential best applications for each of the different modalities.
SESSION 4 : ANALYTIC INNOVATIONS AND REMOTE SENSING
The growing concentration of the global human population into cities has coincided with increasingly rich data from smart environments, social sensing, and the Internet of things. Fusing these data with maps of the built environment and urban vegetation has enormous potential to quantify urban energy and water use, improve urban planning, and target public health initiatives. Currently, fine spatial resolution imagery is prized for mapping urban materials because it can identify object edges. Yet this imagery doesn’t have enough spectral bands to discriminate important differences between urban materials, and it doesn’t provide global coverage. Near-future orbital imaging spectrometer missions could revolutionize our understanding of urban environments, measuring hundreds of reflected wavelengths from the visible through the shortwave infrared. However, these platforms will have pixel sizes > 30 meters, which means we need to develop methods to extract urban surface information at sub-pixel scales. I will present new results in which we used airborne imaging spectrometry to extract fractional estimates of key urban surface classes and urban vegetation condition, obtaining robust estimates across spatial scales. When such data become globally available from satellite platforms, there will be increasing opportunities to produce spatially explicit value-added products for utilities, municipalities, and green technologies.
Use of automated remote sensing techniques to map forest cover is important when modeling environmental quality. However, identifying forest cover with multispectral imagery (MSI) often results in confusion caused by similar spectral profiles between forest and other vegetation. Previous research in forest mapping has included integration of hyperspectral imagery and LiDAR data for tree detection and use of MSI to distinguish tree crowns from non-vegetated features. Since these sources are not widely available to most practitioners, a method was created to discriminate between forest and other land covers using commercial MSI. This research discusses two indices, the Forest Cover Index 1 and Forest Cover Index 2, which were developed to model forest in WorldView-2 satellite imagery of the Beltsville Agricultural Research Center in Maryland. The study site included mixed forest, agriculture, other vegetation, urban features, soil, and water. The tree cover indices exploited the product of either reflectance in red and red edge bands or the product of reflectance in red and near infrared bands. For two classes (trees vs. all other), overall classification accuracy was >85% for the four images that were acquired throughout the year.
This session will look at how Commercial Geiger-Mode technology differs from existing LiDAR systems and what it means to the future of our industry. The presentation will include a technology overview with current and future applications as well as real project examples.
SESSION 1 : ADVANCES IN GEOINT TRADECRAFT - GEOSPATIAL MODELING
Digital Elevation Models (DEM), regardless of origin and resolution, provide information on terrain elevation but applications require transforms to support analyses. Dr. Way has developed a suite of terrain concepts and models that can be applied to a wide range of applications from finding Karez features in the Middle East (utilizing 1 meter data), to quantifying the ability of the terrain to “hide” equipment or personnel. This presentation will summarize the geospatial concepts used to determine relative relief, terrain complexity (texture), terrain slope characterization, terrain slope characterization (uplands/lowlands), and terrain slope characterization (valleys), will be presented. A number of applications of these products will be shown including off-road vehicle mobility, generation of soil mapping, identifying potential sources of construction materials, ambush opportunity, and aerial concealment.
This presentation will discuss recent improvements made to the Monte Carlo Scene (MCScene) code to enable low light situations where the sun is near or below the horizon. MCScene is a high-fidelity model for full optical spectrum (UV through LWIR) hyperspectral image or multispectral image geospatial simulation. MCScene provides an accurate, robust, and efficient means to generate spectral scenes for algorithm validation. MCScene utilizes a Direct Simulation Monte Carlo (DSMC) approach for modeling 3D atmospheric radiative transfer including full treatment of molecular absorption and Rayleigh scattering, aerosol absorption and scattering, and multiple scattering and adjacency effects, as well as scattering from spatially inhomogeneous surfaces, including surface bidirectional reflectance distribution function (BRDF) effects. The model includes treatment of land and ocean surfaces, 3D terrain, 3D surface objects, and effects of finite clouds with surface shadowing. This session will provide a brief overview of how real-time elements are incorporated into the Monte Carlo engine, and will also discuss the recent additional of a polygonal earth cross-section (PEX) model which allows for long atmospheric path simulations such as those found under twilight conditions and highly off-nadir or near horizontal viewing geometries.
SESSION 2 : MACHINE LEARNING AND INDEXING AT SCALE
The evolution of GPU technology has fueled breakthroughs in artificial intelligence. In particular, deep learning has powered innovations in language translation, image search, and driverless cars. Paired with the development of deep learning frameworks and software providers that integrate with NVIDIA’s CUDA platform, the technology has become more accessible to traditional researchers outside the computer science discipline. The result has been massive gains in automation while reducing the compute footprint and infrastructure. At the heart of deep learning approaches is a data-driven methodology to learning the patterns in large corpuses of information. While remote sensing data has always been ‘big data’, the amount of data that analysts and researchers will need to consume will only grow with the introduction of commercial satellite providers. The traditional exploitation algorithms of remote sensing data have relied on engineered, statistical approaches that require domain expertise to implement and deploy. These techniques have proven effective in the environments for which they were designed, however scale, access to domain expertise, and lack of model transferability remain challenges. The integration of deep learning techniques, and GPUs to power them, into GEOINT exploitation offers an exciting and tractable solution to automating some of the most challenging problems facing the community.
Data Integration and reduction is the biggest technical challenge and opportunity of our age. Our approach to consumption and integration is almost entirely manual, forcing people to find, analyze, and correlate data. Unstructured data offers the biggest challenge and must be indexed automatically, at scale, when created, and at machine speeds. Probabilistic data modeling enables us to leverage low-specificity/low-sensitivity indicators to produce high-specificity/high-sensitivity insights by linking large volumes of data from multiple sources. There are five basic steps needed to achieve automated integrated predictive data models: Automatically index the content of all data; correlate data sources; develop models; develop conditional responses to collection, action, and analysis; and, evolve. The tools to do this exist today, are ready for use, and can help us begin to understand and use our data to its fullest potential. The culmination will be entirely new approaches to data.
This presentation will review the application of deep learning technology and techniques to automatically extract objects and features from geospatial intelligence data sources including high-resolution RGB and Pan imagery, high revisit rate imagery, LiDAR point clouds and products, mosaics from drone imagery, and full motion video.
The majority of existing deep learning frameworks (e.g., CAFFE, TensorFlow, NVIDIA Digits) are designed to make training of algorithms fast and efficient when a lot of data and resources are available, but not a lot of effort has been spent on making them work in constrained environments. The purpose of our work is to develop a framework that allows for the easy development and training of deep learning algorithms in low-resource environments (e.g., small-sample datasets).
We will discuss how to use transfer-learning methods for easy training and application of object classification, image captioning, and face detection and recognition algorithms. In addition, we will discuss an architecture based on Docker containers developed at Johns Hopkins University Applied Physics Laboratory that allows users to quickly prototype deep learning algorithms on personal workstations and quickly deploy them into production.
Over 80% of countries do not provide any official reporting of economic activity and multiple reports coming out of some countries, such as China, can sometimes imply conflicting results. This talk will cover how remote sensing and geospatial analytics can fill this gap for industry applications in the financial and energy sectors. Through access, aggregation, and analytics, Ursa is the first to deliver reliable weekly reports of 77% of China’s crude oil inventories. Ursa uses synthetic aperture radar (SAR) to deliver its timely reports and has access to all the world’s commercial SAR satellites. Aggregation of multiple satellites gives Ursa the ability to be SAR sensor-agnostic, creating a virtual constellation with a revisit rate of up to twice a day. Ursa’s image processing experts have developed geospatial analytics to work on multiple formats, multiple incidence angles, and multiple resolutions. Additional, proprietary research is incorporated to provide context, such as tank owner and storage type, for analytics results. Example use cases, such as macro-economic analysis, currency, and stock prediction, as well as other financial applications, will be presented. Data sources that provide reliable global economic intelligence are disrupting traditional energy and commodities paradigms.
The possibility of providing, for any time and location in the world, operational services based on Earth Observation (EO) data very often faces very simple but fundamental issue: data availability in an affordable, timely and consistent manner. The Sentinel-1 constellation represents a solution. First, because SAR technology ensures all-time and all-weather acquisition capabilities, the Sentinal-1 has been designed to guarantee consistent coverage for any location in the world, with an average repeat between 6 and 12 days. Next, Sentinel-1 data are downloadable free of charge via a web interface and API. These data are hence ideal to build services that can be operated not only on-demand scheme, but on a routinely basis, for mapping as well as monitoring. Examples will be shown of different operational applications and services based on Sentinel-1 data, built on enterprise-level COTS software tools interfacing directly with the data archive. Finally, the complementarity of a Sentinel-1 based approach with data from other (very-high resolution) SAR as well as optical missions will also be discussed.
Easy access to a regular supply of free satellite imagery is enabling a range of new services based on monitoring and change detection. We will describe how generic products from Sentinel-2 are being used to plan follow-up actions for more detailed data (satellite and UAV) and on-site inspection in local government and forestry organizations. We will also explain how satellite imagery is made analytics-ready through the creation of Precision Datacubes; stacks of accurately co-registered data provide quality change information with minimal artifacts that allow limited resources to be used in an optimal way by significantly reducing false alarms.
The ENVI Analytics Symposium is being held at the Brown Palace Hotel and Spa in Denver, CO. The Brown Palace is a Forbes 4 Star hotel and is a legend among Downtown Denver hotels.
Guests enjoy access to timeless luxury with a unique sense of place, original experiences and world-class service and amenities. There's simply no better way to experience the Mile High City.
Discounted hotel pricing is available for $209/night (plus tax) for ENVI Analytics Symposium attendees while space is still available. The block of discounted rooms will sell out, so we advise you to book a room as soon as possible.
There are a variety of ways you can get from DIA to the Brown Palace:
Super Shuttle provides transportation from the airport for $25 per person one way and $46 round trip. To make arrangements with Super Shuttle, call 1-800-BLUE-VAN. You can also visit their website: www.supershuttle.com.
A one-way taxi ride from the airport to The Brown Palace is approximately $75.
We would like to thank the following sponsors for their support of the 2017 ENVI Analytics Symposium:
Esri, the leader in geospatial technology, offers imagery tools and content to see the world, find the patterns and share with others. Esri’s ArcGIS platform includes means to manage and serve large collections of imagery for use in ArcGIS and other software such as ENVI. ArcGIS Online provides access to large collections of imagery such as Landsat GLS collection, as well as global elevation datasets.
Sharing data on Amazon Web Services (AWS) makes it accessible to a large and growing community of users who use the AWS cloud for research, new product development, and education. When data is shared in the cloud, anyone can analyze it without having to download it or store it themselves, which lowers the cost of new product development, reduces the time to scientific discovery, and accelerates innovation.
Planet designs, builds and operates the world's most capable constellation of Earth-imaging satellites. Planet's mission is to image the entire Earth, every day, and make global change visible, accessible, and actionable. With the complementary RapidEye constellation, Planet has an image archive from 2009 and a network of more than one hundred partners around the world. See change. Change the world.
Learn more at www.planet.com, or follow us on Twitter (@planetlabs).
NVIDIA's invention of the GPU in 1999 sparked the growth of the PC gaming market, redefined modern computer graphics and revolutionized parallel computing. More recently, GPU deep learning ignited modern AI -- the next era of computing -- with the GPU acting as the brain of computers, robots and self-driving cars that can perceive and understand the world.
DigitalGlobe is a leading global provider of commercial high-resolution earth imagery solutions. Sourced from our advanced satellite constellation, our geospatial offerings support a wide variety of uses within defense and intelligence, civil agencies, mapping and analysis, environmental monitoring, oil and gas exploration, infrastructure management, Internet portals and navigation technology.
The Intelligence Programme Line of Airbus has unrivalled expertise in satellite imagery acquisition, data processing, and dissemination. Airbus provides customized solutions across all markets, and based upon exclusive access to optical and radar satellites, the company delivers an extensive portfolio spanning the entire geo-information value chain.
ASD is a world leader in spectroscopy solutions for remote sensing. Researchers at universities and institutions across the globe trust our portable, rugged and easy-to-use instruments and software solutions for research-critical measurements in the field. To rapidly collect spectra to ground truth hyperspectral and multispectral imaging data, choose ASD.
SARmap's mission is to build and provide an innovative, sophisticated yet simple remote sensing software product, dedicated to the generation of digital information for a better management and risk assessment of Earth's natural/environmental resources.
CloudEO teams with world-leading content and software providers to offer to you a unique geo-infrastructure as a Service bringing together data, software and processing power.
HySpeed Computing’s mission is to evolve the science and business of managing our planet’s resources through development of innovative applications for deriving information from geospatial imagery.