BC: In this episode, we will talk to product line leaders from the Developer and Provider Suite about the latest innovations in ERDAS IMAGINE, GeoMedia, ImageStation, ERDAS APOLLO, and GeoCompressor.
Welcome to HxGN Radio. My name is Brian, and today we are going to discuss the new capabilities of the Developer and Provider Suite related to the Power Portfolio 2018 release. Joining me are Brad Skelton, Product Line Leader for ERDAS IMAGINE; Vince Smith, Product Line Leader for GeoMedia; and Chris Tweedie, Product Line Leader for ERDAS APOLLO. Welcome, gentlemen! Thank you for joining us. So, Brad, let's start with you. Tell us a bit about what the IMAGINE team has been working on.
BS: To begin with IMAGINE, it's a remote sensing product. We've been extracting information from images for quite some time now. For the 2018 release, the IMAGINE team enhanced the product and expanded its capabilities. It's now a full 64-bit application. We've added numerous features to support new sensor models and introduced new capabilities for radar products. But the most significant aspect of IMAGINE is how we deliver these capabilities. In IMAGINE, we have a tool called Spatial Modeler: it allows you to build new algorithms graphically, using individual operators. We've integrated all these capabilities into IMAGINE's operators. In this version, we've added another 90 operators, covering everything from image classification to property modification and extraction, radar and optical photogrammetry, and more. So, we've significantly expanded the suite's capabilities this time around.
BC: Great! There has been a lot of buzz about machine learning. Tell us about machine learning and how it has been incorporated into ERDAS IMAGINE.
BS: One of our main focuses in spatial modeling this time around was in the area of machine learning. It fits well within IMAGINE's realm because image extraction is a form of machine learning. We've had early forms of machine learning in IMAGINE for many years. This area has certainly advanced significantly over recent years, and we have now rolled back some of these technologies into Spatial Modeler. We now have machine learning that can be trained on various image properties and objects, including what's known as deep learning. Deep learning is a technique that uses images and their parts for training to recognize image properties. So, we now have operators that you can train and that will allow you to recognize or find objects in images such as cars, airplanes, or even road and field types. We were really focused on this in the new release.
BC: Very interesting. Excellent. Now, Vince, let's talk about GeoMedia. Tell us about the enhancements this year.
VS: You know, in the middle of last year, we essentially made a roadmap. We visited some customers and saw interest in GeoPackage, the OGC specifications for maintaining self-contained data, and decided to move forward and add full read-write data server capabilities to GeoMedia. This was a kind of pivot in our overall release strategy. But we had great interest from users; many adopted the OGC GeoPackage specifications to work with their geospatial data repositories. Instead of using shapefiles and transferring data back and forth, they use GeoPackage. Furthermore, we had a client interested in implementing some of our mobile capabilities, and our integration required the GeoMedia team to provide the capability to effectively import and export this data into GeoPackage. Such an approach prompted us to add this data server (GeoPackage) to the release. So now we have integration between our line of mobile products and GeoMedia to transfer data between them. People can now export their internal GIS information from the database, place it on tablets, Apple or Android, take it into the field, collect information, and then return and use GeoMedia's data server capabilities to place the data back into their internal corporate environment. We've gone so far as to add GeoPackage as our default storage option in GeoMedia. This is a significant change for us. It was always referenced, but simply because GeoPackage is portable, lightweight, and widely adopted, it made sense for us to give it to our customers as the primary GeoMedia storage option.
Second, of course, building on what Brad just mentioned—spatial modeling and the Spatial Model editor—we've been hard at work adding GeoMedia capabilities as spatial operators, making them available in IMAGINE, and IMAGINE capabilities in GeoMedia. We added things like creating functional attributes, intersection functions, masking, unioning, and they are all now new spatial operators for the 2018 release. We added several new expressions. They are quite important for processing functional attributes. So even with the spatial model editor, you'll get some new expressions that you can use. And, of course, now with GeoMedia, in addition to spatial model operators, you can use the models you work with anywhere in the GeoMedia product. Just as you use queries, functions, categories, and more in GeoMedia, any team in GeoMedia working with functionality can now work on spatial models.
BC: Excellent. What are the new features that have emerged in ImageStation?
VS: Regarding ImageStation, the big news is that we now support many more satellite sensors. Any sensor that can be modeled using RPC (rational polynomial coefficients). We've also added remote sensing modeling. Thus, Spot 7, Spot 6, COMPSAT, RapidEye—sensors that can be used within ImageStation. So yes, this is what we've been working on for the past couple of years and will continue to work on, to add other sensors. But at the same time, we've added support for UAVs to the 2018 release of ImageStation. Not everyone is happy with the low-cost, "black box" approach to generating data from UAVs, and some ImageStation customers came to us and asked us to add the capability to transfer UAV data into ImageStation because they like control. ImageStation gives them control over the quality of the products they produce. And now they can do this with ImageStation's product line.
BC: Alright, thank you, Vince, that's very valuable information! Chris, what's new with ERDAS APOLLO?
CT: Brian, I suppose some context is needed—APOLLO continues to be the enterprise data management platform for many of our clients. And this really means that we provide a technical platform that allows organizations to discover, manage, and deliver their information and other data throughout the organization, both internally and externally. One of the key new features of our 2018 release, launched back in March, has really expanded our path to the implementation of new RESTful APIs. Since this is an enterprise platform, integration with other corporate business systems is obviously very, very important. Starting with our 2016 release last year, we introduced really reworked REST APIs. With the 2018 release in March, we expanded them to cover almost all the workflows that clients would need to integrate our platform into their business system. For example, Clip-Zip-Ship was not previously covered, and now, from 2018, this tool is fully supported and can be integrated.
Another key feature of our 2018 release that I am very proud of is significantly enhanced integration with the Windows security system. In previous releases, we had some level of support for this workflow, but we really rethought it in the 2018 release. This means that everything, from setting up and integrating within domain security settings, has become much, much simpler than in previous releases. From an organizational standpoint, there are direct benefits. We reuse existing domain credentials, as well as roles and security levels, which, for example, are already managed by corporate IT specialists. But from a user perspective, one of the main benefits I like to see is that if you are using a supported browser, you don't have to enter a username or password. With the 2018 release, it automatically logs you in (username, password) if you set it up properly, which is simple, transparent, but, I must add, secure.
BC: Excellent, yes, it's good to know not only about security but also about simplicity, which is nice. Great. Now, what about GeoCompressor?
CT: GeoCompressor is our specialized tool for mosaic creation and compression. Although we have obviously very similar compression and mosaic workflows in Vince's ImageStation tools and Brad's IMAGINE tools, GeoCompressor is a very specialized application that performs only one task, specifically designed for our data provider customers. As part of the Provider suite, we complement data-backed workflows in such a way that we can compress mosaic images much more aggressively than any other application on the market. With GeoCompressor 2018, we have really focused on improving the quality of mosaics and the quality of the spatial management. We have added a number of features that will make it easier for you to work with these large-scale workflows. For example, we have added capabilities for dynamic range adjustment and color balancing. We can also quickly expand the dynamic range, as you can see in any package. Now, all images are stored in a zip file. This is very important. You can open the file directly with your photo viewer, or you can open the file in a supported GIS application. This is very useful for quick validation of your work. It is very important for us to know that you have applied this tool to create, for example, 500 photos of work for planning that could be 100 pages of a project for planning or implementation, you can find in the Viewer application.