We're sorry but this page doesn't work properly without JavaScript enabled. Please enable it to continue.
Feedback

Cloud-Native Asset Model with STAC

00:00

Formal Metadata

Title
Cloud-Native Asset Model with STAC
Title of Series
Number of Parts
156
Author
License
CC Attribution 3.0 Unported:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
Identifiers
Publisher
Release Date
Language

Content Metadata

Subject Area
Genre
Abstract
STAC is a well-known and acknowledged spatiotemporal metadata standard within the community. There are many applications with open-source data; however, there are few adoptions by premium satellite imagery providers. At UP42, we adopted STAC as the core metadata system within our applications that defines how we store data. Last year, we presented how we designed a standard data management system with STAC. As of last year's talk, we developed another catchy concept to standardize data access and processing. We designed a Cloud-Native Asset Model combining existing concepts such as STAC and COG, where we transform all files delivered by providers into a somewhat standard-ish format using GDAL extensively. We want to continue updating the community about our experience and share the takeaways.
Keywords
127
Streaming mediaProcess (computing)Point cloud
Normed vector spaceOpen setGamma functionLaceFile formatExtension (kinesiology)MetadataLibrary catalogDifferent (Kate Ryan album)Computing platformType theoryShape (magazine)MetadataComplex (psychology)Process (computing)AlgorithmData storage deviceMultiplicationSingle-precision floating-point formatExecution unitInternet service providerSatelliteUniform resource locatorInformationStack (abstract data type)Combinational logicSoftware developerLibrary catalogFile formatMereologyProcess modelingCategory of beingComputer fileKey (cryptography)DemosceneLevel (video gaming)Point cloudElectronic visual displayLatent heatMultiplication signDebuggerCASE <Informatik>Computer animationLecture/Conference
Least squaresComputer-assisted translationComputer animation
Transcript: English(auto-generated)
Hey, yeah, I am but so I work at a 42 Sorry for coming up again another collab native, you know like terminology So our problem was basically when you have we are we are we are a platform where you can buy satellite imagery from different providers and if you get one imagery on your storage one
delivery on your storage you end up with many multiple different shapes of Metadata and also the files sometimes they have like they have PNG files TIF files NTF as text files and sometimes they deliver in a format that you really need to
analyze well what's going on with this shapes and Formats and also for the metadata. There are some there are some even people Putting their logo in the delivery. So it's too much information usually so We use stack as a it's it's a very nice solution just to tidy up things
and basically, we we started to Use stack for our storage. So if you go to the storage and maintain this complexity Stack is already tidying up many thing for you and the developers But our problems again
Then there are like many ways of doing stack because stack does not dictate you how you should combine Collections catalogs and items there are very nice. I think definition of item which is more like atomic unit It's a single location single day time, but if you want to combine, you know
how we should do collections and items and we should Name assets keys and properties, etc. You again end up with multiple questions how we should do for Maintaining a storage scene and that was our basically main
Model that whenever you have an asset afford to the asset we call but you can have another asset just one asset that you have on your search you transfer it to your storage and we made it equal to a stack collection
Because sometimes you have stereo imagery or multiple imagery from different part of the earth They can actually relate to different location different daytime so it was a wise idea to have a one step collection to define one asset delivery and one catalog for your auto storage and items
For your like different imagery parts and many many other different stack assets So it's it's looking like this actually Mostly, I'm skipping the properties properties are almost standard in stack or
The assets are usually very different depending on the provider We basically, you know follow the best practices in stack They are pretty useful and most of the times whenever we have a question you just like check the specification and best practices and you find the answer and
What what we do is like we add some same information also for people are using it like we facilitate roles more we try to standardize how we make the titles and we try to also convert every tiff or Every imagery formats to the cloud native assets so that in the later stage you can use it very easily
And this got quite standard I think and it's getting useful. What is useful is that it's very useful for front-end front-end people are very happy with this because they have now reliable metadata where they can get information and easily display on the map and for different
For different providers, it is standard. They don't have to deal with all the complexity They just use this model and what is also like videos for that if you want to process this complexity Depending on all providers have a different shape and if you come up with an algorithm you have to add up for all providers
but in this case, it is also getting much much easy to process all different type of data and It got like it's very getting useful in the processing platform That's it