Author: Peg Shippert
I'm going to talk about someone else's work based on yet another body of work for a minute here, but bear with me because it's worth it. I just ran across an excellent article written by Brad Plumer for Vox based on NASA's Images of Change series. Images of Change is a set of before and after remotely sensed images from various places on the Earth. What I like about Plumer's article is that he includes a wide range of types of change, mostly focused on changes caused directly by human activity. He then describes each of them succinctly and clearly for the lay person. I also appreciate that he includes examples of both destructive (e.g., deforestation) and productive (e.g., reductions in air pollution) changes.
It's well worth a read. Check it out!
The Topaz Solar Farm in California, seen in 2011 (before the farm was built) and 2015. Image courtesty of NASA, Images of Change.
Categories: ENVI Blog | Imagery Speaks
Author: Amanda O'Connor
In the next few years there are some very exciting weather sensors being launched. NOAA’s weather work horse system, GOES, will get a new instrument GOES-R that will greatly improve the temporal, spectral,and spatial resolution of the weather imagery that gets used to make forecasts,watches and warnings. The Advanced Baseline Imager is the primary imaging instrument on GOES-R and built by our parent company, Exelis Inc. There is a similar system called the Advanced Himawari Imager (AHI) on Himawari-8 which is Japan’s weather satellite that launched last year. See the table below forABI’s improvements over the legacy GOES system.
GOES-R ABI facts from http://www.goes-r.gov/spacesegment/abi.html
The Mesoscale mode for imaging is a huge boon to severe weather forecasters. Getting imagery every 30 seconds of a hurricane or tornadic system will provide much greater forecasting detail as well as the ability to forensically model events. See image below showing kinds information more frequent imaging can provide.
In ENVI 5.2 we released a raster series tool that did a great job telling stories about change in imagery. You could animate over one spot or through entire pass of Landsat data. For the casual remote sensing user it got the job done very well. Recently working with a weather customer, we found that animating through GOESimagery at a rate that you would see on the weather report or on a weatherwebpage with GOES imagery at full resolution (~1km pixel for the visiblechannel, 16,000 x 7,000 pixels) was taking a little longer to animate and loadthan we’d like. So our engineering team took the challenge and made some changes. The animation below shows the increased speed.
You are able to annotate and draw ROIs in ENVI while the image is animating, so imagine the type of information you extract while you’re animating—things that are changing. Things where you need the animation to see the change like a smoke plume, water turbitiy, geomorphology, or sea ice movement. Think of classifying features with the added temporal element. World vectors that come with ENVI can be overlaid, but also more local vectors, like home or property locations, so one can see areas being impacted by severe weather. Videos of the animation can be created and shared, all at full resolution. When weather data in the US takes a huge leap forward with GOES-R and ABI, ENVI will be ready to ingest the larger data volumes and provide animation and analysis support tools to go with that data.
Follow me on twitter @asoconnor
Author: Adam O'Connor
The Global Multi-resolution Terrain Elevation Data 2010 (GMTED2010) dataset is the elevation dataset of choice for global and continental scale applications. GMTED2010 is the result of a joint development effort between the U.S. Geological Survey (USGS) and the National Geospatial-Intelligence Agency (NGA). GMTED2010 is a major improvement in consistency and vertical accuracy over GTOPO30.
Starting with version 5.1, the ENVI software installation includes the GMTED2010 dataset for the mean elevation aggregation method at 30 arc-seconds spatial resolution in numerically lossless JPEG2000 format. ENVI provides the GMTED2010 dataset as a convenience for users who need a digital elevation model in their data visualization or image processing workflows. The GMTED2010 dataset can be easily opened in the ENVI software using the "File > Open World Data > Elevation (GMTED2010)" menu option and certain processing tools such as the RPC Orthorectification and Image Registration workflows will automatically use this dataset as the default DEM.
Within the ENVI software installation the GMTED2010 dataset is distributed as a file named "GMTED2010.jp2" that is 152 MB in size and the default path locations are as follows:
Windows: C:\Program Files\Exelis\ENVIxx\data\GMTED2010.jp2
In the next few weeks an alternate version of the GMTED2010 dataset with 7.5 arc-seconds pixel size will be provided on the Downloads section of the Exelis VIS website (user account required). This upgraded "GMTED2010.jp2" file is 1.54 GB in size and provides a higher spatial resolution global elevation model which can improve the accuracy of RPC Orthorectification and automatic tie-point generation in Image Registration (if at least one input image has RPC metadata).
To replace the GMTED2010 dataset in the ENVI software installation with this higher spatial resolution 7.5 arc-seconds version simply overwrite the "GMTED2010.jp2" file in the appropriate path based on computing platform listed above (may require administrative privileges). Here is a screenshot of the 30 arc-seconds dataset distributed as part of the ENVI software installation (left) and the upgraded 7.5 arc-seconds version (right) that you will soon be able to download from our website:
Author: Patrick Collins
* Image used courtesy of Wikimedia
Last week I was speaking to a gentleman at the SPAR International 2015 conference in Houston, Texas who was trying to conduct a series of viewshed analyses using LiDAR to assess how much of an area could be seen from certain vantage points within their area of interest. His problem was that the software package he was using only took into account the Digital Elevation Model, or DEM, as opposed to leveraging the Digital Surface Model, or DSM, that was also available within the point cloud. Basically, when he got to the area where he had conducted his analysis he would find there was a tree, a building, or some object obstructing his view.
What he needed was an algorithm that took into account not only the height of the observer, but the height of their objects around the observer as well. Below is a screenshot of a general viewshed analysis in ENVI LiDAR, which automatically sets the height of the observer to 2 meters above the surface of the Earth, or DEM value.
Notice how the viewshed takes into account the surface of the not only the ground, but of the trees or other objects as well. This is done by building a Digital Surface Model that incorporates all of the points in the cloud to determine what can be seen from a given spot. When an observer point is created, the user also has the option to adjust the height of the observer, to model someone who may be in a tower, up in a tree, or raised above the ground by some other means.
For this point, I've raised the observer up to a height of 17 meters, which results in the viewshed below. Note how the area visible to the observer has improved due to the increased height which allows them to see over the tops of some of the trees in the area.
Aside from height, there are a number of other properties that can be altered to allow for more advanced viewshed analysis. These include the ability to set the radius of the observer's view, the ability to change the color of visible vs. non-visible areas, and the ability to set the azimuth, pitch, and horizontal and vertical field of view. It also contains the ability to set the DSM resolution, and to toggle on and off range rings at a user-specified difference.
A user can also create multiple observers within a scene, which enables the assessment of visible coverage for an area from multiple points. This can be very useful from surveillance or security standpoint. The observer settings can also be modified to show only areas that are visible by all observers, or areas that are visible by any observer.
Finally, all of this information can be transported over to a software such as ENVI or ArcGIS® with a single click for further analysis to calculate information such as area covered or to better understand similarities or differences between observers points.
It's important for geospatial analysis to reflect real world situations. Simple oversights in our understanding of geospatial problems can result in misleading information that can waste the time and resources of the organizations that are using specific solutions in their day-to-day workflows. Spending the time to truly incorporate all of the relevant factors into a solution can go a long way towards increasing the efficiencies of those organizations and teams. After showing this to the gentleman I met at SPAR, he was very excited that he'd never have to go visit a misidentified location again!
*By Virginia State Parks staff (Leesylvania State Park Uploaded by AlbertHerring) [CC BY 2.0 (http://creativecommons.org/licenses/by/2.0)], via Wikimedia Commons
Tags: ENVI, Digital Elevation Model, DEM, Viewshed Analysis, LiDAR, Digital Surface Model, DSM, view, shed, Height Model, 3D
Author: Kevin Wells
I’m really looking forward to the upcoming ENVI Classified User Symposium (ENVIUS), which is taking place on May 14 at the National Geospatial Intelligence Agency (NGA) in Springfield, VA. This is an event that I’ve supported for several years now, and it is fun to see how it has become one of the most anticipated meetings of the year for our customers that work in the Defense and Intelligence community. I regularly have customers tell me that it is a great opportunity for geospatial analysts and scientist to hear and see what thought leaders are doing within the community, as well as an excellent opportunity to network with their colleagues.
This year, Aimee Baldwin has done a great job putting together a full agenda of speakers that are using Exelis VIS technologies to solve some really challenging problems. Some of the highlights of the day include:
- Several discussions on the use of point cloud data from various modalities and sensors for 3D feature extraction and how this analysis can be conducted in a cloud environment using ENVI Services Engine (ESE).
- A speaker from George Mason University will be talking about how he has been able to use ENVI with HSI data for accurate, efficient, large–scale and timely identification and mapping of hydrilla infestation in coastal estuaries and other water bodies
- Representatives from the USGS will be briefing on the research that they conducted by analyzing a wide range of properties from disasters such as World Trade Center, Hurricane Katrina, the Deepwater Horizon Oil Spill as well as various wildfires and landslides. Through this work, they have identified potential safety and environmental issues that may have been overlooked or not identified by traditional remote sensing techniques.
- Work that has been done on the optimization of target detection algorithms so as to suppress noise and cluster anomalies and improve the ability to find specific targets.
- Image scientists from NGA will be discussing a new methodology of using the VIIRS Day/Night Band (DNB) for nighttime light analysis. They have then used ENVI Services Engine (ESE) to more efficiently process their data on a server and are executing the same processes for data residing on the desktop. Both of these implementations use a multiprocessing approach, which is greatly improving their efficiency and saving significant amounts of time.
We are also going to have our Enterprise Product Manager, Mark Bowersox, provide a view through the looking glass as to what is on the horizon with the various VIS products.
Keep in mind that this event is only open to US Citizens with a TS/SCI level security clearance.
If you have any questions about it, please do not hesitate to reach out to myself (email@example.com) or to Aimee Baldwin (firstname.lastname@example.org).