Summer downtime? It certainly doesn't feel that way over here at UP42. But with the summer drawing to a close, we wanted to share what we've been up to during this August to iteratively make accessing, purchasing, and analyzing satellite imagery easier.
Check out everything that we've been up to this month 👇
——— ✨ What's New? 🚀 ———
Complexity is one of the biggest impediments to successfully accessing and analyzing satellite imagery. With these UX updates, we’ve removed redundancy, streamlined the process, and future-proofed our platform to enable us to offer you even more data sets in the future. Let’s take a look at the changes that are helping us deliver on our vision.
To make an informed decision on what satellite image you want to purchase, you need to access information from delivery timeframe, cloud cover, and resolution to, sometimes, more technical information, such as scene ID and incidence angle.
Whatever information you need to make the right purchasing decision, you can now access that information in a single spot within your regular search, discover, and purchase workflow. Just click on an image within the catalog search results and select the “Image Information” tab to get everything you need.
So you’ve found that perfect image for your use case. The cloud cover is low enough. The delivery estimate is in line with your expectations. And the scene area fully covers your AOI. But how can you be absolutely sure?
You can now see the scene preview right in the purchasing flow. As you click into the image thumbnail, exactly where you can view the image information, you can click into the scene to see a preview and take that extra step of validation before hitting the “Buy” button.
Once you’ve chosen and previewed your scene, you can choose to purchase or process the image, or indeed both. If you intend to use the image for one-time processing within UP42, you should select the processing option, where you will later choose a workflow to process the image within.
If you intend to use the image multiple times or want to download and take the image outside of the UP42 infrastructure, you should choose the purchasing option. The image will then be delivered to your storage, where you can download the image or copy the asset ID for us in a processing workflow. The next update showcases how these images can be used in a processing workflow!
As some of our future data sources will require additional compliance steps, accepting the End-User License Agreement (EULA) and requesting access to restricted blocks will also take place within this popover window, unifying the experience for all users across all future catalog search data sources.
So you’ve purchased a Pleiades or SPOT image from us. What now? With this new block developed by UP42, you can now analyze data from your storage right within the UP42 platform. This new block, the beta launch of this functionality, allows you to use an UP42 workflow to process SPOT or Pleiades data from your storage with any compatible processing block.
This functionality is available via the Python SDK, API, or console. Here, we’ll focus on console usage but you can see how to use the block with the SDK or API in our documentation. However, regardless of whether you are processing data via the SDK, API, or console, you’ll still need to copy your Asset ID from your UP42 storage.
The flow of using this block is relatively similar to using any data block on UP42. Build an UP42 workflow using the Processing From Storage block as the data block followed by any compatible processing block. In this example, we are simply converting data from my storage, the output of a Pleiades download block, from DIMAP to GeoTIFF.
As we move from the workflow builder to setting the parameters in the job configuration screen, we simply need to paste the Asset ID into the job parameters enclosed in [“asset ID”], select “Live Job”, and if necessary, adjust the parameters of your processing block(s). And it’s as simple as that. Review your price estimate and run the job. Of course, as you have already purchased this data, you will not be charged again. You’ll only be charged for the costs of the processing blocks and infrastructure costs.
This block version of our processing from storage functionality is a beta feature. As a result, we recommend usage from experienced technical users only and primarily through our Python SDK or API.
We will be working on streamlining the experience and widening the scope of the feature to enable the re-processing of more than just Pleiades and SPOT imagery in the near future. So keep your eyes peeled in the coming weeks and months.
You may remember Spacept from our recent vegetation management webinar. A leader in the use of artificial intelligence to analyze satellite imagery and provide actionable insights, they’re back with a brand new block on the UP42 marketplace.
This new building detection block, which processes SPOT or Pleiades imagery, has been trained on tens of thousands of semantically annotated images from a range of countries in order to identify buildings with an accuracy of over 80%.
The Spacept team has put together a tutorial to show you how easy it is to detect buildings using their block on the UP42 platform.
Our latest partnership with HyperVerge, one of India’s leading AI companies, enables the detection of small structural changes to homes and properties for local government tax assessment and code enforcement. The block also helps monitor legal and unauthorized construction of roads, buildings, and artificial islands in remote regions by federal or defense agencies and keeps GIS maps updated in rapidly growing urban areas.
The algorithm processes Pleiades or SPOT imagery and is optimized for a high recall score. This means that it generally accurately identifies all areas of infrastructure change but may also return false positives. As a result, it is particularly useful as a precursor to manual QA or as an input for more focused analysis.
——— 📅 In Case You Missed It 😱 ———
Hear from industry experts and see technical walkthroughs into how satellite-based analysis can augment existing drone or helicopter-based insights, drastically reducing costs for vegetation management.
19.08.2021 | Jasmine Fleming | 2.37 billion people are without food or unable to regularly eat a healthy, balanced diet… Read more
That's it for now. But we have tons of exciting things in the works that we'll announce in the coming weeks and months, so keep your eyes peeled in Release Notes, or if you can wait a whole month, we'll meet you back here in September!