QGIS 3D, point cloud and elevation data
This is a modal window.
Das Video konnte nicht geladen werden, da entweder ein Server- oder Netzwerkfehler auftrat oder das Format nicht unterstützt wird.
Formale Metadaten
| Titel |
| |
| Serientitel | ||
| Anzahl der Teile | 266 | |
| Autor | ||
| 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. | |
| Identifikatoren | 10.5446/66523 (DOI) | |
| Herausgeber | ||
| Erscheinungsjahr | ||
| Sprache |
Inhaltliche Metadaten
| Fachgebiet | ||
| Genre | ||
| Abstract |
|
FOSS4G Prizren Kosovo 202381 / 266
10
17
23
44
45
46
47
48
49
50
52
53
80
84
85
91
110
116
129
148
164
167
169
173
174
181
182
183
186
187
199
202
204
206
209
215
241
248
265
00:00
Güte der AnpassungKontextbezogenes SystemBitCoxeter-GruppeComputeranimationVorlesung/Konferenz
00:27
Mobiles InternetSynchronisierungQuellcodeHumanoider RoboterEDV-BeratungDienst <Informatik>Migration <Informatik>Hill-DifferentialgleichungEDV-BeratungSoftwareentwicklerSpeicherabzugVorlesung/KonferenzBesprechung/Interview
00:42
EDV-BeratungMobiles InternetHumanoider RoboterSynchronisierungQuellcodeMigration <Informatik>Dienst <Informatik>SoftwareentwicklerPunktwolkeInklusion <Mathematik>App <Programm>sinc-FunktionMereologieComputeranimationVorlesung/KonferenzBesprechung/Interview
01:10
PunktwolkePunktEDV-BeratungOpen SourcePunktwolkeZeitzoneRückkopplungVollständiger VerbandRechter WinkelTaskOverlay-NetzBitmap-GraphikVektorraumOffene MengeKartesische KoordinatenWeb ServicesGenerizitätComputeranimationVorlesung/KonferenzBesprechung/Interview
01:57
RückkopplungBenutzerbeteiligungOverlay-NetzGenerizitätOpen SourceVorlesung/Konferenz
02:24
PunktwolkePunktFramework <Informatik>Prozess <Informatik>RückkopplungProgrammbibliothekVisualisierungVorlesung/KonferenzBesprechung/Interview
02:58
PunktwolkePunktFramework <Informatik>Prozess <Informatik>EDV-BeratungSoundverarbeitungFokalpunktVorlesung/KonferenzBesprechung/InterviewComputeranimation
03:14
Objektorientierte ProgrammiersprachePunktwolkeFramework <Informatik>Vorlesung/Konferenz
03:48
PunktwolkeVorlesung/Konferenz
04:04
PunktwolkeCASE <Informatik>Güte der AnpassungDifferenteFlächeninhaltKategorie <Mathematik>PunktVorlesung/Konferenz
04:28
InterpolationPunktwolkeComputeranimation
04:58
RenderingDateiformatPunktEDV-BeratungElektronische PublikationMultiplikationsoperatorAutomatische IndexierungDemoszene <Programmierung>PunktwolkeVisualisierungDateiformatProzess <Informatik>FlächeninhaltComputeranimation
05:36
SymboltabellePunktWasserdampftafelGebäude <Mathematik>GeräuschFlächentheorieVektorraumPunktwolkeTypentheorieDifferenteLastComputeranimation
05:57
Attributierte GrammatikProzess <Informatik>Klasse <Mathematik>ATMSpannweite <Stochastik>LastVolumenvisualisierungSpieltheorieKoordinatenPolygonnetzDatenbankVektorraumInterpolationVolumenvisualisierungComputeranimation
06:14
VolumenvisualisierungRenderingATMMeterKreisflächeDatentypBrowserOrakel <Informatik>Gewöhnliche DifferentialgleichungAdditionComputeranimation
06:32
AbschattungVisualisierung
06:53
TopologieIdentifizierbarkeitComputeranimation
07:09
EDV-BeratungDefaultMinimumMinimumGebäude <Mathematik>Ordnung <Mathematik>SkalarproduktMereologiePunktMathematikComputeranimation
07:34
Turm <Mathematik>DatentransferBridge <Kommunikationstechnik>WasserdampftafelGebäude <Mathematik>DatensichtgerätKlasse <Mathematik>PunktEDV-BeratungService providerRegulärer Ausdruck <Textverarbeitung>DigitalfilterLokales MinimumZahlenbereichLastZoomPunktwolkeGeräuschMaßstabQuellcodeMathematikAttributierte GrammatikEinfache GenauigkeitRenderingDreieckVolumenvisualisierungPunktwolkeDemoszene <Programmierung>VolumenvisualisierungTextur-MappingEinsTriangulierungSchnittmengeResultanteKlasse <Mathematik>VektorraumPunktDreieckFlächentheorieDifferenteComputeranimation
09:12
ZoomPunktwolkeEDV-BeratungBildschirmmaskeElektronische PublikationDigitalfilterQuellcodeMathematikReservierungssystem <Warteschlangentheorie>Shape <Informatik>PunktwolkeClientAttributierte GrammatikElektronische PublikationComputeranimation
09:53
PunktwolkeEDV-BeratungServerProzess <Informatik>SoftwareClientIndexberechnungPunktPunktwolkeGlobale OptimierungElektronische PublikationInformationsspeicherungClientURLStreaming <Kommunikationstechnik>Automatische IndexierungSchnittmengeSoftwareStandardabweichungJSONComputeranimationVorlesung/KonferenzBesprechung/Interview
10:45
PunktwolkePunktSoftwareClientIndexberechnungProzess <Informatik>ServerDatentypFiletransferprotokollQuellcodeProtokoll <Datenverarbeitungssystem>Elektronische PublikationAuthentifikationKonfigurationsraumDatenbankEDV-BeratungDateiformatKartesische KoordinatenElektronische PublikationClientService providerFlächeninhaltProzess <Informatik>Vorlesung/KonferenzComputeranimation
11:25
BrowserEDV-BeratungServerComputeranimation
11:40
Prozess <Informatik>PunktwolkeSchreiben <Datenverarbeitung>Lesen <Datenverarbeitung>ProgrammbibliothekPunktVorlesung/KonferenzBesprechung/Interview
11:56
EDV-BeratungPunktwolkeTreiber <Programm>DateiformatParallele SchnittstelleFilter <Stochastik>Prozess <Informatik>PunktwolkePunktParallelrechnerComputeranimation
12:42
FunktionalVorlesung/Konferenz
12:57
BenutzerfreundlichkeitParallele SchnittstelleRandwertFunktion <Mathematik>Bitmap-GraphikEin-AusgabeEDV-BeratungVollständiger VerbandEin-AusgabeArithmetischer AusdruckPunktwolkePunktZahlenbereichParallele SchnittstelleNabel <Mathematik>Radikal <Mathematik>Dichte <Physik>CASE <Informatik>SkriptspracheVersionsverwaltungGebäude <Mathematik>p-BlockRandwertProzess <Informatik>Mailing-ListeBefehlsprozessorPolygonFramework <Informatik>Filter <Stochastik>TopologieParallelrechnerComputeranimation
14:47
EDV-BeratungAlgorithmusVektorraumPunktDateiformatUmsetzung <Informatik>AxonometrieInformationEndliche ModelltheoriePunktwolkePunktPlastikkarteVersionsverwaltungComputeranimation
15:11
RechenwerkAchtTesselationMeterVektorraumVorlesung/KonferenzBesprechung/Interview
15:31
AlgorithmusDateiformatInformationVektorraumPunktAxonometriePunktwolkeStellenringPhysikalisches SystemVirtuelle RealitätÄhnlichkeitsgeometriePunktProjektive EbeneUmsetzung <Informatik>PunktwolkeVirtualisierungVorlesung/Konferenz
15:46
Elektronische PublikationPunktwolkeBitmap-GraphikDateiformatVirtualisierungVorlesung/Konferenz
16:04
PunktwolkeEDV-BeratungDateiformatPhysikalisches SystemStellenringÄhnlichkeitsgeometrieVirtuelle RealitätPunktElektronische PublikationKeller <Informatik>RPCStellenringSoftwareViewerVisualisierungPunktPunktwolkeComputeranimationVorlesung/Konferenz
16:33
PunktwolkeService providerRenderingProzess <Informatik>AlgorithmusEDV-BeratungPunktwolkeVirtualisierungFlächeninhaltProzess <Informatik>Quick-SortComputeranimation
16:57
EmulationEDV-BeratungPunktwolkePunktVisualisierungComputeranimation
17:12
PunktEDV-BeratungSichtenkonzeptKartesische KoordinatenSpannweite <Stochastik>ZoomKontrollstrukturMultiplikationsoperatorComputeranimation
17:28
EDV-BeratungDiskrete-Elemente-MethodeMaßstabAuflösung <Mathematik>Maß <Mathematik>Bitmap-GraphikPunktwolkeFlächentheoriePi <Zahl>PunktEinflussgrößeFormation <Mathematik>SchnelltasteEinflussgrößeObjekt <Kategorie>Gebäude <Mathematik>Demoszene <Programmierung>Computeranimation
17:57
EDV-BeratungPunktwolkePunktAttributierte GrammatikVolumenvisualisierungComputeranimationVorlesung/Konferenz
18:20
FlächentheoriePunktVolumenvisualisierungVariablePunktwolkeAlgorithmusProzess <Informatik>EDV-BeratungPolygonnetzPunktwolkeMapping <Computergraphik>TesselationFlächentheorieTextur-MappingComputeranimation
18:36
XINGRauschenTypentheorieVorlesung/KonferenzBesprechung/Interview
18:53
FlächentheoriePunktVolumenvisualisierungVariableAlgorithmusProzess <Informatik>PunktwolkeEDV-BeratungPolygonnetzSchreiben <Datenverarbeitung>OrdnungsreduktionFramework <Informatik>Kartesische KoordinatenComputeranimation
Transkript: Englisch(automatisch erzeugt)
00:08
Hello and good morning, everyone. Welcome to this presentation. In the next half an hour, I'll walk you through some of the features in QGIS 3D.
00:20
The exciting bits already court has covered, but I try to add some more context. I'm Sabir from Lutra consulting. three years and some of the features you work with, we have
00:43
developed. So if it works well, it's us. If it's crashing and gives you problems, it's also us. Blame us. We also have a mobile app, which my colleague will talk after that, based on QGIS. Okay.
01:08
So QGIS 3D has started a crowdfunding campaign to include point clouds or LIDAR data inside QGIS. We teamed up with two other open source companies, North Road and
01:21
Hobo, to do this task. The main problem was really finding the right time zone in Europe, North America and Australia to have a meeting to talk, but eventually we got there. The whole initiative, the whole idea
01:41
behind that crowdfunding was based on feedbacks we had from the community. This LIDAR data was lacking in QGIS. There were some third parties and some tools that did conversions, but there was no native tool in QGIS. Also, there were some open source tools
02:03
to open LIDAR data, but on the other hand, you couldn't overlay it with the rest of your GIS data like vector, raster, web services, so we thought it would be good that QGIS, which is a very generic desktop application, to
02:22
have this feature. We prepared a proposal and put it to the community, and they generously funded our work. The first one was to integrate the visualisation, the PDAL library.
02:43
It was in 2020 and we successfully raised enough fund to do that. Then again, we had lots of feedback from community to add more tools like elevation tools and some more visual effects when they
03:02
load data, and we had another crowdfunding, €81,000, and that one was successful as well. Then the last one, which I'll focus on, it's the introduction of PointCloud's processing framework in QGIS, which was
03:22
done late last year, October, November, and now, as of the new release of QGIS 3.3.2, it should be available. Oops, I broke the whole thing. What did I do?
03:54
Last tab.
04:02
PointCloud data are okay, good. These are some examples of point cloud data, in this case, you will see over a billion points covering a
04:21
large area and with different categories, classifications, and then another example is the mobile scanning data, this is from Vienna, and last example, it's
04:44
from Vienna, it's a large point cloud covering Sydney, I think. These are all visualised in QGIS, and to get there, things we had to overcome was handling large amount of data.
05:05
The common file format for PointCloud is last, these are very good for storing data, but they are not optimised for visualisation and processing, so we did behind the scenes some indexing, when you load
05:22
the data for the first time, it does some indexing and creates some files, then it reads those more efficiently and you can visualise large areas. Then point clouds have got lots of data
05:42
behind it, in this example, you have classifications, different land use. You can easily, in QGIS, load the data and show those classifications. Another type of data, it's the Z, you can have the symbology, the renderer set based on
06:04
your elevation data. So you can create some impression of your digital elevation model and then some of them have got RGB and you can see it similar to aerial
06:21
photography. Also, in addition to 2D, you should be able to visualise that all in 3D as well. Then to improve the visualisation, you have also the dome lighting, so you can
06:40
easily identify the features. This is without and the other side is with. You can see easily what features you have. Another one, both iDoming works in 2D and 3D.
07:02
Then ambient occlusion, you can easily identify trees, for example, by having this feature. Then ordering of your points, in this example, if you look at this part, for example,
07:23
you can't, the black dots now reordered and you will see proper roof of your building. And the order change from bottom to top. The other improvement, as Kurt mentioned, we have now classifications plus the percentage of
07:42
the points, so you can see for each class what percentage of points you have. Also, you can see the non-standard classes here as well. Filtering, if you have a large point cloud and you want to see a
08:01
specific filter, a specific class, similar to vector data in QGIS, you can apply a filter and do it based on classes or based on elevations, et cetera. You can easily filter those and once you apply it, your map will just show the result of the
08:22
filter. Another useful tool is to have this 2D renderer to set the scene in 2D. When you set the scene in 2D, it used to be the fact that when you go to 3D, you have to apply all the settings to the 3D, but by having this,
08:42
they both show the same thing in different map. Another one showing the point clouds as surface, it fills the holes. Also, one of the challenges were the size of the triangles, if you have very steep
09:02
triangles, you can filter those and it skips the triangulations for those odd ones. Export to file, everybody loves shape files, so why not exporting point cloud to shape file and
09:25
You can export to both export and filter the data on the fly and also set the attributes you want to export and there are some tools that unfortunately don't support point cloud directly and
09:43
the clients want the point cloud in DXF or shape file, this is a way to convert them. And then cloud optimised point cloud. As I mentioned, last files are good for storage
10:04
of data, but they are not very optimised for displaying and processing large data sets. So there is a standard COPC and There are a lot of things similar to geotiff,
10:22
when you create a cloud optimise, it creates an index and you can easily either stream the data over a URL or you can just have it locally, then it will be much easier for your client to visualise or process the data.
10:48
APNIC, so you can use APNIC 6 to opt for this format. This is an example of using it as a file or you can just point it to your EPT or COPC and if
11:04
you are a data provider, you want to cover a large area by streaming the data to your user's desktop. Also, it will be fast to process
11:23
from the client side as well. So this is an example of LIDAR data. This is EPT, the predecessor to COPC and it's easily visible, can be visualised in QGIS.
11:43
Point cloud processing, we based the processing on PDAL, PDAL is similar to GDAL library but for point cloud, it handles read and write of data. Also, it has some processing tools but it's
12:02
slightly different to GDAL, instead of having a point cloud, and in standalone processing, you have filters and it has got a Lego approach, so you need to create pipelines by putting those filters together and to create a pipeline, for example, to create a DTM out of your point cloud.
12:23
It gives you ultimate flexibility but to the parallelisation is done by user, within the scripting, but for desktop user, again,
12:44
it will be a hurdle. What we did, we wrapped up all those functionality in a new tool called PDAL wrench and it does create a
13:05
a point cloud to a boundary and you can run it either from terminal, your shell or it's integrated within QGIS as a processing tool and you can run it from there. It also automatically
13:21
handles the multi-processing, the parallelisation based on the number of CPUs you have, so you don't need to do any manual work on So by doing that, we have got this list of processing tools in QGIS in 3.32.
13:45
It's the starting point. It might not look very comprehensive, but it's the building block that you can in future add more and more filters and pipelines from PDAL so that the framework is there.
14:01
I'm going to show you a few examples of some of them. This is the clip by polygon, so you create a, you have got a point cloud input layer and a polygon and run it and it produces the clipped version. Density, you can create a density of your point cloud where you have got high
14:22
number of points and low numbers and then filtering the data, so you can filter the input data based on expression similar to the one I mentioned earlier and this one creates boundaries, so for example
14:43
you can create boundaries of your buildings, trees and then the popular one is to export it to raster, so you can create elevation models out of your point cloud data.
15:00
There are some others like thinning, so if you have a very large coverage of the point cloud and you want to have a resampled version to give an overview, you can create a thin version of that. You can export that as a vector layer.
15:20
You can merge data and also tile data, so if you have a large coverage, again, you want to have you want to have them in 100 metre by 100 metre or kilometre by kilometre, you can do that. Fixed projection, conversion, they are self explanatory and the more exciting one is
15:42
virtual point clouds, it's when you have lots of large data similar to GDAL virtual raster, that with the last files and then the format of it
16:03
is based on stack API, so each of those will become an item collection. You can, if you have a stack compliance software, you should be able to do it similar to the one to cop C viewer and it can reference local files and remote files. the point clouds, then you can load it in QGIS,
16:25
visualise all the coverage of your point cloud, as you zoom in, the point clouds will appear. Also the beauty of this is you can do all sorts of processing I mentioned beforehand. Let's say you want to clip the data or create DTM of large
16:49
data, you can do it in a specific area and then just select an area and clip the data or create DTM in a specific area. So this visualisation is also available in 3D, this
17:03
example with 37 billion and it's smooth because the data is streamed and you can easily see the data. it in 3D. I'm running out of time, so I quickly go through the 3D improvement, Kurt has mentioned most of them,
17:23
but one that is available now. It's measurements in 3.32, global, okay. This one is 3D measurement, so you can smartly move
17:43
around and measure buildings in 3D scenes. It nicely snaps in on your objects and you have measurements in X, Y and also Z. I give you a preview of some of the upcoming
18:02
features, confirmed works in QGIS 3.3. In QGIS 3.6, we will introduce point size renderer, so you will be able to change the size of point cloud based on some attributes.
18:21
You can also see the point clouds as a surface in 2D maps in QGIS 3. 3. 3. 3.34 later this year, we will have support for 3D tiles that's funded by CZM, so you can create those nice Google Earth type of things inside QGIS.
18:42
Also, if you want to bring across some of those interesting Pdol pipelines for applications and we should be able to introduce that
19:00
easily, so come and talk to us.