Merken
Exposing NASA's Earth Observations
Automatisierte Medienanalyse
Diese automatischen Videoanalysen setzt das TIB|AV-Portal ein:
Szenenerkennung — Shot Boundary Detection segmentiert das Video anhand von Bildmerkmalen. Ein daraus erzeugtes visuelles Inhaltsverzeichnis gibt einen schnellen Überblick über den Inhalt des Videos und bietet einen zielgenauen Zugriff.
Texterkennung – Intelligent Character Recognition erfasst, indexiert und macht geschriebene Sprache (zum Beispiel Text auf Folien) durchsuchbar.
Spracherkennung – Speech to Text notiert die gesprochene Sprache im Video in Form eines Transkripts, das durchsuchbar ist.
Bilderkennung – Visual Concept Detection indexiert das Bewegtbild mit fachspezifischen und fächerübergreifenden visuellen Konzepten (zum Beispiel Landschaft, Fassadendetail, technische Zeichnung, Computeranimation oder Vorlesung).
Verschlagwortung – Named Entity Recognition beschreibt die einzelnen Videosegmente mit semantisch verknüpften Sachbegriffen. Synonyme oder Unterbegriffe von eingegebenen Suchbegriffen können dadurch automatisch mitgesucht werden, was die Treffermenge erweitert.
Erkannte Entitäten
Sprachtranskript
00:00
trying bowler presented might begin on behalf of this team at NASA type this is a collaboration between NASA's Goddard Space Flight Center in a GPL and that's the least we don't know the mapping company that develops it allows you the this conference so now retired but today is opened you data that we have open services that we now have and so open software that we have to handle these things so the start of thought about the spacecraft that we have the observations that they make so it all
00:37
started back in 1960 with Tyrus 1 and this is ours conception of it majestically soaring through space over hurricane and with what came back imagery was well it like that
00:52
so but but the thing is is the 1st time that but we realize that you can use satellites to look at the earth and effectively for whether so it was a great starting point and we followed up with a name a series of
01:05
spacecraft just large and 64 is 9 hours conception we the data back and actually
01:11
looks a little more scientific we've got some some scale bars there and some things look like an audit track and and yet people
01:20
like this who actually operate ground stations so you know making progress look off into the machine is so if a sawtooth
01:30
today and you got that we got about 15 or 16 of operational satellites in orbit and this these are pretty different than that the commercial variety that you might see and you know I am first-level envious of some of things a planet the suing hirer spacecraft from but these guys couple of scientific domains and actually in the details of of what companies they can see so so 1st of
01:58
sample image of show years from are you 1 spacecraft to this the volcano in Indonesia last year essentially volcano image that means there's new before and after so here's the before
02:08
and after so this line and quite a big impact on erupted earlier this year and for the between the 2 here
02:15
that's 30 meter for picked 13 meter per pixel but resolution history this was
02:24
from motor on the terror spacecraft that was snowfall from the east coast earlier this year and no this is this a pretty picture you're reflectance but we have put courage reflectance it uses the a red-green-blue channels of this this but we can also build scientific products in the other bands in this case notice has 30 K channels I believe a different wavelengths from infrared to visible so you decide to do things like quantify
02:52
how much snow actually fell was the snow coverage map here now and
02:59
intelligent picks up the snow pixels mostly and those they got the clouds and so he decides to do some sort quantitative analysis here using the metadata of likewise this
03:10
is a different kind of measure of water here using notice again this is 10 15 intra for pixel it in California followed this year this is a measure of vegetation and and as anybody lives in West coast knows California not doing very well this year for for vegetation water
03:30
so I I'll talk about about using these observations so we got a long
03:36
history of of of of making is observations on 1 1 of these instruments have global coverage to them uh that there's a ton of scientific products that are available so that to the settings is a lot of diversity and richness I inner now you
03:55
know were in the modern era these days you don't have to you know use these these all reels things the things are online system goal major data in some sense but as he must have worked with what they know
04:08
and you know you don't usually have yes in problems like data formats are very different from processing levels the resolutions of course corpora geospatial data recorded systems are never never so to but also this is this is a problem that we have an acid is that uh a lot of our datasets are time-varying so that something is not very well supported in OGC or across a lot of a lot of tools so all address that will but it's picture of ours most that kind of data as well so every day we get about a half termites In our total or comes about 10 petabytes now the but what we wanna do is to be able to open the state up for for outsiders to use of the instead of having to spend too your career grad school learning how to use this data should be a little bit a little bit easier use we so um this solution voter but
05:09
we're doing right now so am we're trying to make
05:14
things visual so where basically we want we want to users interact the data visually and to discover data visually and to that end we we have a set of services that are that a applied open access so there's no no registration key just use it is it tiled datasets of metallic imagery to the the datasets and and uh it's open to use for mapping clients forward for G. Del scripts and GIS clients it is the 2nd we also have an open source a browser-based client worldview which Michael talked about here shortly so some background on how a doing this that use
05:57
nested things are locked up in these binary granules is that on the moment of mouse so that 1 little chunk use considered a granule and book where I work with the data providers to basically projector that that's data apply would've Robinson due if I quality control and the nearest residents to some some and but new repeated again for the next renewal and the next rental then you you have a whole day's worth of these granules
06:27
so you can do that as sample and into these full resolution daily global mosaics when you do it again for the next day and the next day
06:37
when you do that again for some uh
06:40
for different products in this case is the sea surface In the sea surface temperature product of bands and we've got about
06:51
uh 70 80 90 products a so far right now they're in In gives that that show a combination of well of action of March of the and
07:02
so I knew it was you have all a lot of this imagery then we need a way to ingest it 1st and to that end we have introduced system called the entire the image exchange only open-sourced next year really tell storage format and a tile serving the server does or a matter after format in honor of respectively the subdural cover secular today he wouldn't details of knows about open source and available and get out and I in the optically that the storage and serving of software that's been around for for quite a while adjust open-source recently and it's so the uh global atomic in details this is called the individual toxin themselves but but our goal here is that high-performance Rasta file server that's what this is cheese so this is what you have
07:56
that imagery and uh as tiles answered out to a clients you can access it through GDL scripts and also suggest clients but there's a caveat on the GIS kinds and that not all of them support time-varying data sets so that's really big challenge for us and work with OTC due to handle the uh time a little bit more robustly right now it's so a challenge so um so the WTS slum et al services
08:28
our primary spec without camel and telling w as well and it uses the sample sample tower request that's a RESTful access where you got and the product names a time projection and you're zoom level in rows columns and you get back my style the Canary Islands instance but
08:54
so the types of edges that were severely right now we've got a reflectance kind of product so that others natural color of the upper left is a false color there's a couple of false color of income nations in the lower left there and on the on the right is cut service besides parameter can renderings where and in this case the snow cover is in blue and the sea ice in future the and then the clouds are appear the the of baseline we
09:25
got 40 from that projections of a sports O star of a geographic we will be extended to the polls and then liberated we will also have the Web Mercator projection so compatible with of the commercial sources of is in terms of the products we
09:42
have in there now really a motor shop at at this moment there so some couple other products near to there's analysts and straight attention good resolutions for document years this is 250 meters and and coarser In some cases here and the future will have aster a couple and products will be known to 30 meters the right now we have about the last 2 years of imagery available and were working with notice to reprocess everything from from history so that about the year 2000 and so what those 10 15 years for
10:18
pixel like so this is going to get to your so this is a clear day of services go and from lesson to but so you have this basically globally twice per day 2 different motor centers it's also available within neural time
10:40
so within 3 to 5 hours of being observed that you can pull down and duty so that helps us do things like Applied sciences like looking at floods I'll get wildfires tuition shipping so
10:55
here's a five-day period here groups
10:59
with looking at uh wildfire
11:03
upset assembly from last year
11:07
of and then the the whole here and this
11:10
is near-normal asking this spring looking at sea ice so from a plan when to get your the this daily images of that your plan when you get a icebreaker through through so thing note that a so we currently have from Michael shows more samples of of the current literary have become the in the future
11:32
or working our really excited about the idea being swimming pp imagery in which a lot like notice what modest doesn't have as this day night band so it's my band that what you see the imagery at nite from the city so you get a lot of city lights you get some fishing vessels you can see I get some fires that are burning and so the exciting season now over delta you
12:01
there's a sets from product and working on called weld basically takes the uh cloud-free imagery and on a weekly monthly and annual basis the some of the product adamantly monthly from 2011 and we can do is take the best the products and then make use traditional products as well so this point where N is a weak help because we've got a lot of Indra the remembering in and not enough people to work with so the news I take a quick detour and so we
12:36
have open position begins so that you like working with satellite imagery you like to operate the NIC systems on you always figure out what to do next I would full-time position guided a tendency of all of the colonies or the matching so
12:57
that program here only to another people using this and EPA has
13:02
this massive they created with which shows the the ground stations which measure air quality here as these point sources NIF underlaid as a base layer and 1 the daily um uh modes symmetry sir from Gibson any can see that the some smoke here and so provides in some context for the for the ground stations but the quizzes and
13:26
its product that measures aerosols in near and and they is this for for context as well got DB is integrated
13:36
with us as well so you can do some you're visual storytelling but with our dreams
13:42
and box e-mail as very lightly 1 we can said we was great entire archive and how we do that without you know really knowing does and that they didn't they did a great job of that what they did is basically take all of our our quality imaging upper-left sorted it to to pick the best cleanest ones and and then merge them on the best pixel basis you have their their call this atlas which my levels are our from from injury and trunk horribly oversimplifying what they're doing here but you can get the idea of is also being used in a museum setting as
14:19
well so here's a signs on a sphere of of 5 or 6 foot diameter sphere that shows entry point over time but really well poses you know
14:33
these are open services so we want the community to build a whatever they want whatever whatever's meets their needs with our drinks with because really opened things up quite a bit we have some of light examples and some source code using bare-bones down the clients in leaflet layers in Google Maps Bing Maps and that means you can build in building it find that way and so I think there is the kind of room for new educational apps science arts decision-making apps mobile apps surfing apps you know the the ocean temperature is in dreaming about going somewhere where it's warm so take effect is so I'll pause here actually turn over to my who's been
15:22
thought about our flight that we're building with with using Gibbs called world of you over
15:32
to international sliding what the of the
15:41
the it's an army would like to give a on with a wide
15:46
demo of this but the internet section at working out all that
15:48
well here today so I'm screwed were Henshaw video instead of a this application is available online iterative a massive echo of an economic Chapelle later if you like this is most of bombs events like right now so this is actually a volcanic eruption in Iceland of and on in the walking away as in if I did a would be more pronounced it anyway so now been see here that the and stories and here the know that 1 up in combination with motors in it actually this we see the while the trail there the animal authoring
16:35
considering movies he prepared I guess so
16:40
we this quick here now this is actually the true colors that you what you see on you'd see with like your normal i from together to the group from common out but when yet she had this additional band combinations yet you get more detail more information from the surface the most of the
16:56
brightness temperature product we can put that in catchy
16:59
see as well that the big anomaly
17:02
there in the temperature the intersection happened for
17:08
couple days so this was actually of word and you know that I get a new back couple days and the house evolve over
17:18
time some basic cladding he the through over these are
17:21
good tha
17:30
security have some virus this was
17:32
on get if I see with the mouse had
17:45
I guess it is worsened virus braille and sometime in August this is like the border between the you know you're working on and you can see here on the right up here because they on because you're such a huge cloud there the came from a and I should friend was in the area at the time he said the was pretty intense we can you Maloney get sees through at the
18:13
extent of the wall fires the happened there at the
18:15
time I Ch
18:25
case so this is actually what this view this video what topic around lunch time so you can see this is like showing on the 1 day basically there was a lot that want passing the way that we have so as you unitary pops up is immediately available so for viewing here also have these were tracks enables again through the words that you had so what kind of a a service was actually a able visualize idea real-time imagery as a king then so some of these you do with the application that we have built using Gibbs
19:09
on which 1 more popular features exodus
19:11
people do a image downward thing Mr. just pick a region they go like I can choose what resolution you want you want something that's a completely like large image can pick a different of file types I the download button in Algeria
19:26
image 40 and then you can say that but you local this
19:30
chain
19:36
is right so before it supportive of projections so this is looking at the bottom in also the
19:44
Antarctic it is here that you big hole here
19:47
was actually there isn't any they like down there to get you go back for times for see how
19:53
the whole changes depending on the season this was a
20:01
fun things that have no temporal component here we have
20:06
like a bell 50 layers of cells in your right now on the c to search left the clear that you want and appears these a fires certainly bring down in Africa mostly for vegetation of forever you know clearing agricultural land he put multiple layers along with many layers you want sort filtering by areas of interest as well on was change a word you want the way is of the regular map index you change the color palettes if the it starts interfering with each
20:43
other you get their view of it it will have the ability
20:50
to on it like she Gallo the underlying data security wikified is that she sees something like the 1 I downloaded election quite a pretty simple interface freaking discomfort and the green it 1 downward it pick 1 they want so if a deal is simply click on the download button
21:09
lecture give you what was the of length of reconnection team that data tha
21:20
get 1 share what you see you know and cut a prominent and send it through with
21:28
her whatever In as lasting ready don't some ordering images wealth to succeed in
21:33
very upright that she shows us the
21:35
eyes on here at the polls against foresee how the extent of the series changes over time of caesar like but the variable of all we have done the world view of what we said that this actually gives an open source that can be used by anybody the build
21:53
you know whatever quite that they want what they have existing client to can use In corporate that data inside their stuff so that any questions the
22:09
only some and this gives severed
22:17
metadata service from laughing able to you display way now
22:22
when no new notice images available at the extent of
22:26
inventory I mean you mean which some class I give you the question of the the
22:32
extensive of what's available gives and
22:38
I'm asking if there's a metadata service so to query what the dates of available images all their extends
22:47
and so there is a part of the WMT aspect there's a get capabilities request that you can make it'll tell you the start and end times that are available will surpass to global so there's no I usually no global are no the regional that extent that it'll give you horrendous to broaden the question I guess out is your search service available through some rest some point and can I get there from the search results rather than just having it in the spring was that for the UN and the actual on search for the data for download so on the such acquiring the echo cattle get that information so you know what you have to all the history of the a oneself and you yeah well and is the thing with that is that since you can use agony he's re verb and you know Brecht friend we implement it you know the thought the focus of our application but we actually provide a simple way we can actually get which like their self the it's
23:58
worth there was a subset search available this year so when they reflect a certain Reynolds down here to only the who did so quick question you mention you have a you will have the future plans for obvious standard DNI band imagery and will happening planting comparator history cut the MST Elias images work which would DM
24:31
alignments is there there part and that can be less data project sure and that's not right now we don't that would be nice that for continuity right because it's the only other my time product available now unfortunately right now yeah guess from 1998 to 2012 so visit and music in here and makes the suggestion but the plans right now out there but they were questions the answers the bigram with
25:09
few
00:00
Web Services
Mapping <Computergraphik>
Offene Menge
Software
Kollaboration <Informatik>
Web Services
Software
Datentyp
Taylor-Reihe
Luenberger-Beobachter
Sigma-Algebra
Computeranimation
00:36
Satellitensystem
Punkt
Raum-Zeit
01:03
Virtuelle Maschine
Zentrische Streckung
Weg <Topologie>
Arithmetische Folge
Kippschwingung
Nichtlinearer Operator
01:29
Satellitensystem
Nichtlinearer Operator
Domain-Name
Landau-Theorie
Stichprobenumfang
Orbit <Mathematik>
Vorlesung/Konferenz
Wahrscheinlichkeitsnetz
E-Mail
Bildgebendes Verfahren
Varietät <Mathematik>
02:08
Spiegelung <Mathematik>
Pixel
Gruppe <Mathematik>
Meter
Materiewelle
Biprodukt
Gerade
Computeranimation
Bildauflösung
02:51
Mapping <Computergraphik>
Metadaten
Subtraktion
Pixel
Wasserdampftafel
Einflussgröße
Streuungsdiagramm
Quick-Sort
Computeranimation
03:30
Domain <Netzwerk>
Menge
Luenberger-Beobachter
Biprodukt
Biprodukt
Menge
Analysis
Computeranimation
03:54
Abstimmung <Frequenz>
Bit
Total <Mathematik>
Prozess <Informatik>
Atomarität <Informatik>
Adressraum
Bildauflösung
Übergang
Physikalisches System
Dateiformat
Analysis
Computeranimation
Entscheidungstheorie
Spezifisches Volumen
Übergang
Gradient
Systemprogrammierung
Dateiformat
Biprodukt
Bildauflösung
Aggregatzustand
05:06
Mapping <Computergraphik>
Client
Web Services
Menge
Registrierung <Bildverarbeitung>
Open Source
Web-Applikation
Visualisierung
Skript <Programm>
Visuelles System
Schlüsselverwaltung
Computeranimation
05:54
Web Services
Spezialrechner
Erneuerungstheorie
Momentenproblem
Mosaicing <Bildverarbeitung>
Mosaicing <Bildverarbeitung>
Gamecontroller
Bildauflösung
Beamer
Service provider
Computeranimation
Bildauflösung
06:34
Web Services
Spezialrechner
Flächentheorie
Mosaicing <Bildverarbeitung>
Gruppe <Mathematik>
Gruppenoperation
Schaltnetz
Bildauflösung
Biprodukt
Computeranimation
07:02
Web Services
Bit
Open Source
Bildauflösung
Physikalisches System
Computeranimation
Überlagerung <Mathematik>
Client
Web Services
Software
Parkettierung
Server
Skript <Programm>
Dateiformat
Speicher <Informatik>
Datei-Server
Bildgebendes Verfahren
08:26
Web Services
Parametersystem
Spiegelung <Mathematik>
Natürliche Zahl
Zoom
Biprodukt
Computeranimation
Übergang
Datensatz
Web Services
Rechter Winkel
Datentyp
Stichprobenumfang
Volumenvisualisierung
Projektive Ebene
Turm <Mathematik>
Kantenfärbung
Streuungsdiagramm
Instantiierung
09:24
Web Services
Momentenproblem
Open Source
Axonometrie
Polare
Güte der Anpassung
Biprodukt
Term
Computeranimation
Benutzerbeteiligung
Rechter Winkel
Meter
Projektive Ebene
Gerade
Bildauflösung
10:17
Web Services
DoS-Attacke
Subtraktion
Pixel
Web Services
Datenmanagement
10:54
Web Services
Quelle <Physik>
Automatische Handlungsplanung
Gruppenkeim
Datenmanagement
Reelle Zahl
Frequenz
Bildgebendes Verfahren
Computeranimation
11:30
Punkt
Menge
Gruppe <Mathematik>
Basisvektor
Biprodukt
12:35
Ortsoperator
Systemprogrammierung
Offene Menge
Satellitensystem
Ortsoperator
Physikalisches System
Optimierung
Computeranimation
Satellitensystem
13:00
ATM
Punkt
Symmetrie
Open Source
Biprodukt
Kontextbezogenes System
Computeranimation
13:33
Pixel
Quader
Prozess <Informatik>
Basisvektor
Systemaufruf
E-Mail
Packprogramm
Bildgebendes Verfahren
Computeranimation
Übergang
14:17
Soundverarbeitung
Kugel
App <Programm>
Client
Bit
Web Services
Punkt
Kugel
Durchmesser
Vorzeichen <Mathematik>
Dämon <Informatik>
Quellcode
Computeranimation
15:21
Open Source
Gebäude <Mathematik>
Client
Vorlesung/Konferenz
Visualisierung
Computeranimation
15:43
Autorisierung
Weg <Topologie>
Demo <Programm>
Hydrostatischer Antrieb
Schaltnetz
Iteration
Kartesische Koordinaten
Garbentheorie
Ereignishorizont
Computeranimation
Videokonferenz
Internetworking
16:29
Addition
Flächentheorie
Gruppe <Mathematik>
Schaltnetz
Information
Kantenfärbung
16:54
Einheit <Mathematik>
Wort <Informatik>
Biprodukt
Programmfehler
17:16
Computervirus
Computersicherheit
Computeranimation
17:36
Computervirus
Flächeninhalt
Rechter Winkel
Streuungsdiagramm
18:12
Weg <Topologie>
Sichtenkonzept
Web Services
Echtzeitsystem
Visualisierung
Wort <Informatik>
Kartesische Koordinaten
Maßerweiterung
Videokonferenz
19:08
Dateiformat
Bildgebendes Verfahren
Bildauflösung
19:30
Verkettung <Informatik>
Minimum
Projektive Ebene
19:51
Mapping <Computergraphik>
Multiplikation
Flächeninhalt
Regulärer Graph
Automatische Indexierung
Rechter Winkel
Mathematisierung
Zellularer Automat
Wort <Informatik>
Zusammenhängender Graph
Kantenfärbung
Quick-Sort
Computeranimation
20:40
Sichtenkonzept
Computersicherheit
Schnittstelle
21:07
Dicke
Gemeinsamer Speicher
Temperaturstrahlung
21:27
Variable
Sichtenkonzept
Open Source
Mathematisierung
Anwendungssoftware
Reihe
Maßerweiterung
Bildgebendes Verfahren
Computeranimation
21:52
Client
Computeranimation
22:12
Metadaten
Web Services
Klasse <Mathematik>
Besprechung/Interview
Vorlesung/Konferenz
Bildgebendes Verfahren
22:33
Quelle <Physik>
Resultante
Metadaten
Punkt
Web Services
Mereologie
Vorlesung/Konferenz
Kartesische Koordinaten
Information
Fokalpunkt
Bildgebendes Verfahren
23:56
Teilmenge
Rechter Winkel
Gruppe <Mathematik>
Automatische Handlungsplanung
Vorlesung/Konferenz
Projektive Ebene
Paarvergleich
Biprodukt
Analytische Fortsetzung
Bildgebendes Verfahren
Standardabweichung
Metadaten
Formale Metadaten
Titel | Exposing NASA's Earth Observations |
Serientitel | FOSS4G 2014 Portland |
Autor |
Boller, Ryan
|
Lizenz |
CC-Namensnennung 3.0 Deutschland: Sie dürfen das Werk bzw. den Inhalt zu jedem legalen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen. |
DOI | 10.5446/31714 |
Herausgeber | FOSS4G, Open Source Geospatial Foundation (OSGeo) |
Erscheinungsjahr | 2014 |
Sprache | Englisch |
Produzent |
Foss4G
|
Produktionsjahr | 2014 |
Produktionsort | Portland, Oregon, United States of America |
Inhaltliche Metadaten
Fachgebiet | Informatik |
Abstract | The satellites which comprise NASA's Earth Observing System (EOS) have a long history of capturing rich datasets with global coverage over extended periods of time. While the data itself is rich (and open!), it can be a daunting task for uninitiated users to find suitable datasets, learn the data format, and subsequently find interesting phenomena. Even for those who are familiar with the data, it can be a time consuming process. But thanks to the proliferation and maturity of open source geospatial software, NASA has been able to build an imagery ingest pipeline, open source tiled imagery server, and open source, web-based mapping client to encourage exploration and discovery of NASA datasets. This talk will describe how NASA is building these capabilities through the Global Imagery Browse Services (GIBS) and Worldview client, demonstrate how others are building upon them, and show what it takes to integrate NASA imagery into clients using the GIBS API. |
Schlagwörter |
data visualization satellite imagery tiled mapping web services |
Ähnliche Filme
