ViRGiS - GIS in Virtual Reality
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Title of Series | ||
| Number of Parts | 237 | |
| Author | ||
| Contributors | ||
| 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 | 10.5446/58682 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
|
FOSS4G Argentina 2021155 / 237
12
15
16
23
26
36
44
52
53
54
59
72
90
99
114
121
122
123
124
127
129
130
139
154
155
166
203
204
219
223
224
230
00:00
Open sourceComputing platformPoint (geometry)Client (computing)Data structureGroup actionActive contour modelCross section (physics)Level (video gaming)OvalWater vaporEndliche ModelltheorieSoftwareProduct (business)1 (number)Projective planeVideoconferencingCore dumpVirtual realityProfil (magazine)File formatDimensional analysisTouchscreenKernel (computing)Object (grammar)Workstation <Musikinstrument>CyberspaceObservational studyDisk read-and-write headSet (mathematics)Population densityVisualization (computer graphics)Presentation of a groupMathematical analysisBuildingUniform resource locatorSoftware industryUser interfacePolygon meshScaling (geometry)Type theoryFlow separationAsynchronous Transfer ModeInteractive televisionMultiplication signResultantFitness functionGraph coloringConfiguration spaceRaster graphicsTerm (mathematics)Computer fileLine (geometry)Vector spaceDesign of experimentsDirection (geometry)Query languageRow (database)Windows RegistryInstance (computer science)Three-dimensional spacePoint cloudCartesian coordinate systemInterpreter (computing)Archaeological field surveySpektrum <Mathematik>Dependent and independent variablesReal-time operating systemImplementationGeometryLibrary (computing)Functional (mathematics)INTEGRALFunction (mathematics)SurfaceProxy serverTexture mappingVolumeShape (magazine)Moment (mathematics)QuicksortMereologyWebsitePlanningTrailElectronic visual displayFamilyUniverse (mathematics)Default (computer science)Physical systemSystem callUniformer RaumPlug-in (computing)CodeMechanism designWindowKeilförmige AnordnungDivisorFile viewerLink (knot theory)Software developerSpeech synthesisSinc functionFrequencySpecial unitary groupFreezingRight angleGoogolDigital rights managementCharacteristic polynomialSlide ruleDifferent (Kate Ryan album)Goodness of fitKey (cryptography)Computer architectureSheaf (mathematics)View (database)Data conversionPreprocessorAbstractionGreatest elementDiskreter LogarithmusMetropolitan area networkClosed setOcean currentComputer-assisted translationDiagramProgram flowchartComputer animationMeeting/Interview
02:39
TouchscreenComputer animation
03:03
CASE <Informatik>SoftwareComputing platformTouchscreenOpen sourceVirtual realitySoftware developerDigital rights managementProjective planeMultiplication signComputer animation
03:41
Information technology consultingSoftwareCASE <Informatik>Sample (statistics)VideoconferencingPoint cloudSoftwareVideoconferencingPoint (geometry)Presentation of a groupCartesian coordinate systemObservational studyType theoryComputer animation
04:40
Population densityKernel (computing)Workstation <Musikinstrument>VoxelSoftwareHausdorff dimensionIntegrated development environmentComputer-generated imagerySimulationTouchscreenWordReal numberVirtual realityPoint cloudPopulation densityFunctional (mathematics)Kernel (computing)Dimensional analysisThree-dimensional spaceFunction (mathematics)WebsiteCyberspacePlanningWordVisualization (computer graphics)Multiplication signPresentation of a groupProxy serverTouchscreenMathematical analysisVolumeQuicksortProjective planePoint (geometry)Data conversionMechanism designMoment (mathematics)Workstation <Musikinstrument>FamilyComputer animation
09:12
SoftwareCore dumpProduct (business)Open sourceFile format1 (number)Data analysisLeast squaresR-ParitätFlow separationIndependence (probability theory)Data modelVirtual realityComputing platformUser interfaceDivisorReal-time operating systemFlow separationMereologyThree-dimensional spaceFile formatEndliche ModelltheorieObject (grammar)Raster graphicsMultiplication signFrequencyOpen sourceCyberspaceBuildingVisualization (computer graphics)Core dumpPolygon meshKey (cryptography)Sheaf (mathematics)Dimensional analysisLibrary (computing)Characteristic polynomialFunctional (mathematics)TrailDisk read-and-write headCartesian coordinate systemData structureVector spaceMoment (mathematics)Instance (computer science)Extension (kinesiology)CodeQuicksortINTEGRALProjective planeProduct (business)Computer animation
15:27
File formatData modelIndependence (probability theory)Core dumpFlow separationSoftware1 (number)Object (grammar)Information technology consulting1 (number)Client (computing)GeometryWindowComputing platformLibrary (computing)File formatCore dumpThree-dimensional spaceMoment (mathematics)Shape (magazine)Cartesian coordinate systemOpen sourceImplementationUser interfaceVideoconferencingFunction (mathematics)Texture mappingGraph coloringMereologyINTEGRALCodeDisk read-and-write headType theoryElectronic visual displayTouchscreenDefault (computer science)Windows RegistryProjective planeView (database)Universe (mathematics)Different (Kate Ryan album)Endliche ModelltheorieTerm (mathematics)Set (mathematics)Row (database)Open setAbstractionProduct (business)FreewarePolygon meshInstance (computer science)Object (grammar)Virtual realityCyberspacePlug-in (computing)File viewerComputer animation
21:22
SoftwareTouchscreenVideoconferencingComputer animation
21:45
SoftwareQuery languageDesign of experimentsLine (geometry)Observational studyWater vaporConfiguration spaceTerm (mathematics)Computer animation
22:34
SoftwareLine (geometry)Archaeological field surveyDesign of experimentsLevel (video gaming)Dependent and independent variablesInterpreter (computing)Cross section (physics)Raster graphicsOvalProfil (magazine)Water vapor
23:35
Gamma functionUsabilityMaxima and minimaSoftwareComputer wormSurfaceAnalog-to-digital converterActive contour modelPoint (geometry)Data structureProfil (magazine)Cross section (physics)Group actionDimensional analysisTouchscreenSurfaceComputer animation
24:46
Programmable read-only memorySoftwareWorld Wide Web ConsortiumComputer reservations systemPolygon meshTouchscreenScaling (geometry)ResultantGreatest elementGraph coloringCartesian coordinate systemReal-time operating systemMultiplication signCross section (physics)BijectionProfil (magazine)Computer fileLink (knot theory)Uniform resource locatorSet (mathematics)Type theoryComputer animation
26:53
SoftwareData model1 (number)SoftwareComputer fileMultiplication signVector spaceAsynchronous Transfer ModeRaster graphicsPolygon meshOnline chatInteractive televisionMetropolitan area networkComputer animation
28:36
DiagramProgram flowchart
Transcript: English(auto-generated)
00:14
Hey, I could be talking to you Give me now I can hear you. Yeah. Yeah, everybody's been telling me not to ad-lib
00:27
Apparently I can't make the banners work some ad-libbing all the way through the time and everybody's Yeah, yeah Steven it's a it's a it's a skill when I get scared I freeze up and I don't move
00:44
Steven Rammage said he thought I was frozen I thought you were looking for your cat. Yeah, everything's going great on this end. We're
01:00
With Andrea hasn't fired me Andrea Andrea, I'm mispronouncing the names. I am I can get Paul I can get that right so So I've done one thing right I've got the right name hey You're halfway there and where where are you located? I am currently in central London
01:25
excellent and Couldn't be further from the foot if I tried Current temperature Pretty close to freezing. It's a really horrible day. Oh, man We had ice storms. Oh, man. Holy crud
01:42
Well, welcome to a London September Goes from blazing Sun to ice storms before, you know, I like it here in southeast Tennessee. We've got clouds and It's kind of it's okay. Nothing. Nothing terrible. So I guess we'll get started
02:03
Since I'm probably running late now Yeah Andre keeps telling me I'm not fired from my speaking gig. Who knows so ladies and gentlemen We have Paul Harwood and we're we're going to talk about Virgis virtual GIS and virtual reality
02:22
Harwood has been in software industry for 30 years with roles and cap Gemini Nokia and Google covering development architecture project management sales and product development So with that I will dive off and add in your screen And good luck. Okay
02:42
Can you see my screen? I? Can Okay Okay, two screens. Sorry. That's what the problem is. Can you see my screen now? There's a good engine. Can you see the slides? You are good
03:03
Okay That's I've got another screen. Okay, we're halfway there. Yeah, my name's My name is Paul Howard Randy just said I'm basically a long time
03:22
Software developer and project manager And I'm presenting some work my company and a associated company fund which is called Virgis Which is a open source tool platform for doing GIS in virtual reality
03:40
Um So the partners are runet software, which is basically a small software consulting company run by me does GIS particularly with the kind of background in archaeology and and and and historical types of applications based in London and Geo consulting company called ISCOID based in Swansea in Wales
04:02
Which I don't know why I just put on the really naff Welsh accent when I said that And we basically have been working together on this I'm gonna go as quickly through YVR, some background into Virgis. I've got a couple of example videos because unlike Christopher I did not read the presenter guidelines and therefore I don't know I'm not supposed to do it
04:22
and the main case study, which is a Quarry case study done by ISCOID. The abstract did talk about a point A point cloud case study as well. We didn't get the permission to use that data I've got a video of point clouds in Virgis
04:41
Instead So at this point, I'd usually in this presentation go into what GIS is and why we need GIS And why we need three dimensional GIS. I'm actually assuming that this audience doesn't need any of that So i'm going to go straight into Why virtual reality? And what does virtual reality bring to us?
05:00
And i'm going to start this conversation this off by by just going through two Exemplar use cases which kind of drove our thinking about why to do this One from kind of an archaeological background and one just talking about, you know, geophysics This is Josh Emmett. This is not to say that Josh Emmett in any way
05:23
Who's a doctor in archaeology in the University of Auckland in any way associated with the project This was just a paper that he presented that kind of sparked off some of the thoughts And this is a typical use case. I think Christopher mentioned it as well About going through an archaeological trench Where increasingly these days they was called every single find using a total station. So they gave a huge
05:45
Three-dimensional three-dimensional A huge point cloud. All point clouds are three-dimensional Of data points. What in this case they'd like to do or they did is run a A kernel density analysis over the points to create a probability function
06:03
a volumetric probability function Which then because the Z dimension is in archaeology a proxy for time dimension gives you a time function for the probability of particular activities across the
06:20
The plan of the site The problem they've got is there's no really useful way of actually visually inspecting and understanding the volumetric output from the point cloud kernel density analysis Basically because all you get is this vaguely woolly thing on the screen
06:42
I don't know if you kind of represented it in the corner down there So one this sort of how can we understand and get into the data? Uh, well, especially when it's three-dimensional and even more when it's volumetric um and the other case which is Much more of the driving case
07:00
Which is the the geophysics? understanding and Basically, um geophysicists like the people at scoid, which is Run by my brother. So I will occasionally talk about my brother unconsciously um the people at this guy They they get remote sensing data from under the ground and they put that together into a three-dimensional model and try and understand it
07:25
And that's a complicated thing for the practitioners to do um Because And I think actually i'm going to nick Something that christopher said in one and a half presentation two presentations ago
07:42
uh and the typical Gis and data visualization visualization tools that I use at the moment you're outside the data looking in you're above the data Trying to understand this thing that's happening on the screen And you're trying to sort of see it from outside of the data um
08:01
Whereas what the particular practitioners really need and also wants To be is actually inside the data surrounded by the data understanding the data And interacting with the data in a naturalistic and intuitive word And this is what virtual reality brings you you can Create a true three-dimensional space with data in it. I mean true three-dimensional
08:24
um, you can then enter into the data enter into that space and enter inside the data and Understand it using The The three-dimensional senses that we have evolved and we have grown up with
08:41
And then uh with verges at the moment you can interact with the data in Um mechanical and understandable and intuitive ways Move the data around And so the virtual reality gives you is this um Being immersed in the data and when we've we've worked with practitioners
09:01
For the first time looking at this the most common thing that people say is wow. I'm actually inside my data uh, this is the reaction that we We try and create from people So that's just a quick understanding about what we're trying to do in virtual reality and why virtual reality would be different to not virtual reality, um
09:24
I'm now going to go through a kind of a quick overview of uh of this platform that we've called verges so verges is a It's a it's an open source platform Um To allow um gis to be done within uh virtual reality
09:46
um We created this as part of our work to understand, you know, because because it's going came to us and said Uh that they wanted to do data visualization of gis within virtual reality So we programmed it in such a way to create an open source platform
10:03
The other that if other people wanted to use they could use um The key factors about what verges is i'm going to start off by saying what it isn't Verges is not intended to be a tool for creating visualizations of um
10:23
It's not called create a tool of creating Bottle uh visualizations of things of pictures within reality it's not Like city jason a way of uh, you know creating a virtual virtual reality model of a
10:41
City, it's not like some people are doing within archaeology a way of creating a model of what a particular Um building would have looked like at a particular period It is a model for for for visualizing and viewing data um sometimes those things might be indistinguishable, but a lot of time they're not I if you want to see a
11:04
raster within three dimensions with some Cross sections in the art sections within three dimensions or co-located and being Tided and create a mesh that identifies particular features within that data. You're not in any way creating anything that that is
11:22
photorealistic or realistic But but you are interacting with data in virtual reality in 3d and that is the model that verges tries to do Uh, uh, it's other it's other key factor is it's um intended to be used in real time with native data formats to allow the data to be
11:42
edited and then to store the within the native format, so it's designed it's intended to To to to look at data in native formats and not accept the data to be pre-processed It is as say a an open source community-led project Uh, at least the core is i'll come back to that
12:03
The core is the the part of the product that takes the data and creates objects within a three-dimensional space With certain characteristics and and and and visualizations um That's in then intended to be used by other applications, uh, which and they need to add certain things
12:24
So i'll come to in a minute So we came up with some key design criteria, um, the key because the key key The key key design criteria was no pre-processing Uh, no need to take it and this is not in any way intended to
12:42
Say that other people are wrong by doing this but our model is we want to take the native data We don't want to have to take that through blender create objs create meshes and structures which we then Bake into a a custom VR application We are looking to create an application that can in real time take data in its usual format and then create
13:07
visualizations in virtual reality that that that that that Show that data Uh, the other thing is the key thing is that the user should be able to edit data um
13:20
We're getting there Um, we can for instance at the moment edit meshes Dxfs we can edit vector data editing things like rasters we then get into some of the The contextual models about what you actually want to do there so we can move them around and we could geolocate them
13:40
But we can't we can't edit the data and i'm not sure we'll ever get to that point or want to um We wanted a Separation between the core functions which I just mentioned the ability to Represent and indeed edit um data within the virtual reality space from the ui and the headset and the
14:01
integration between the ui and the head The u between the user and the u and the headset um This is Primarily for two reasons for for two reasons, um One is that there are many different headsets and they're moving fast and trying to create a an open source project that could
14:25
Keep track of the headset market Uh would be very very difficult So obviously separating the two functions allows the library to exist without having needing to have knowledge of the headset and then allows people to take that library and and integrate it with
14:40
The headset of their choice and that creates more flexibility Uh, and the other one is to allow innovation in the user interface and user business oriented user interfaces and user interfaces in the sort of use case that we're talking about Uh is a very new and leading edge space. So, um Nobody really knows how to do it. Well, so the
15:02
Scope for innovation that basically means that that we can see which way it goes. Um We also wanted the code to be extensible Uh, so that other people could could use it to represent their own data in the any way they want to Uh, we've done this using an entity model
15:21
Uh, the the verges platform itself is a very permissive mit license So anybody can can take it and use it in any way they want to and The entity model is represented by a schema, but the schema is extensible basically You know, you can create your own schema and plug it in in replacement to the the default one Which allows new types of layer and new types of model new types of entity to be added to verges without changing the core code
15:47
which is actually what we do for the The the the the lead project product which is a product that we call it when we're calling verges geo Which has the the basic verges plug-in Um the basic verges library and adds a couple of layers particularly around boreholes which are hybrid layers
16:05
Which take data from all the other layers and represent them in particular ways Um, it is also os independent. Uh, it runs currently on windows on Mac and on linux obviously all the the vr stuff and the but the headset is primarily on windows
16:23
that's because that's the market at the moment, so the the linux and Uh a mac is a desktop type viewer and in fact we have For all platforms for windows as well. We have two Two clients at the moment A desktop client and a vr client
16:42
I will be showing some videos in a minute. Uh hopefully and the videos are all taken from the desktop clients because Taking a video from a vr from a headset client is actually really difficult mostly because people don't keep their heads still It's very unnatural to keep your head still when you're looking through when you're working when you're walking through data
17:03
Uh, but if you take a video of that you get something that makes the people watching the video c-sick so we will be showing something that comes from the desktop clients uh, which is more like a desktop viewer, uh, it's the difference is the difference between walking through data or flying a drone through the data and
17:23
so you'll see it's more like somebody flying a drone through a data and giving you a a television screen view of what they see the technology background, um Burgess is based on unity uh Primarily because unity is 50% of the market and it works and it's available
17:43
um And it allowed us to kick start the project really quickly Um unity of course is not itself open source. Uh, it is widely available and freely available especially for startups and if unity allows you to compile a client and then you can license the client any way you want to
18:04
uh, but if obviously if your revenue gets above hundred thousand uh dollars a year Then you have to start paying something back to unity that's the the the The compromise that we've taken at the moment. I think somebody's probably going to ask about Godot Uh, and at the moment we are
18:20
To use the original term origin of the name waiting for Godot um Because Burgess is written in in unity or indeed it was written in Godot as well. It's written in c-sharp. Um It's written in c-sharp mono. So it is that that's it that self is open source as well It's built upon gdal pdal and mdal as the main as the as the data abstraction layers
18:43
um, we make a lot of use of mdal probably more than than than qg does for instance because uh meshes are Significantly more important in three-dimensional. Um, uh gis we've also got some other, uh, uh,
19:06
Free and open source c-sharp libraries that you're being used in there. There's one called geometry three Sharp, which is the basic is the the core Geometry tool that we use, uh net dxf because we load dxfs in native
19:21
Format not through gdal, uh, and it's all reactive based There's a there's actually a reactive implementation for unity that we use Um, as I say the core implements the the entity model the visible objects So it's basically placing the data within the three-dimensional space the virtual reality space
19:41
And any consumer applications of the of the library then have to provide the user interface the lights and the cameras So way virtual reality works is there is a coordinate space this cartesian coordinate space as christopher mentioned as well Um, we place objects within that The objects have shape they have color they have texture
20:03
Um, but to see them you then have to replace a camera within the space and it's the output from the camera either gets passed to the the screens on your Head-up display or the screen on the pc And you have to move that camera around within the space to represent the movement of the person
20:21
All of that integration is is not part of the verges library. That's part of the consuming application Uh, finally verges library itself is published as a upm package upm as probably most people on this do not know Uh is the unity packet management system. It's basically npm
20:40
Uh They've just created their own A small universe of npm and like mbme scoped scoped registries and there's one scope registry Which is totally open and there is one by an open source community and that's open upm Which is the one that verges is published on so That's the technology
21:01
Oh Very late the entity model. Uh, it's a very typical one It is a layered entity bottle the record set and there is similar to a layer in verges uh, and then anything that gdol og rp del m del or or indeed an obj or 3ds can be loaded as a feature collection
21:24
um, i'm just going to quickly show some videos probably not as many as as I was intending or maybe even less because Because the screen is not obeying me
21:40
No, i'm afraid we're going to have to go without the videos, but i'm going to go to Uh a use case and this is a query Within darbyshire in the uk called dalo Um where they wanted a particular particular study the study is of a feature which is called a doe line
22:03
um, which is a northern english term For a sinkhole. So this is a place where water disappears into the ground the study was intended to understand the the underground um configuration of the rock particularly within with uh
22:25
Intention to pump a large amount of water into the Into basically into the fracture and hope it would go away. It's a way of getting rid of the water from the quarry um Topography is it's a a depression in the ground um
22:40
The geophysics they took was they took a 12 resistivity cross sections This is basically a tool they use to understand the resistance of the ground underneath the the cross section line Uh, they took 13 ground gpr ground penetration radar profiles again. This is just
23:00
Firing a radar signal into the ground and collecting the responses to understand the different layers under the ground and Electromagnetic magnetic survey across the top which created a raster um then the interpretation is you put all these three together when you try and look for the The doe line structure, uh under the ground and trace where that goes to understand the amount of
23:24
The amount of water that it could cope with Um, it says the open voids appear as bodies of extremely high resistance, you know, uh, high water levels are low resistance Um, I just very quickly I'll go through what would be the traditional workflow and then talk about the way that this is different within vr
23:42
So you can understand The the benefits that it brings so traditionally What they would do would be to go through the the resistive cross sections, uh, try and identify uh points on them that would represent a
24:00
Resistivity Um Um Contour or resistivity contour and then once they've done all of this they use some tools to try and create a three or two point five dimensional surface from those points and that would then represent A a transition within the rock structure and they take volumes out of that
24:24
Obviously because they're just creating points creating the the work of creating the points from the cross cross sections and working through a two-dimensional screen to to actually do that is Kind of highly skilled and there's been and and and takes a practitioner that's thinking in in three dimensions and seeing it too
24:40
And then because they then have to automatically create the profile the profile is not optimal So the way that this was done within within verges was uh, the data was all collected as a set of layers Uh, the layers were entered through the the desktop tool which we call landscape. Um, just just collects the the
25:02
The date the the the the location of the data the link to the date of the file name basically, uh, the symbology the the the Crs And puts those into a adjacent schema, which is then picked up by by the the the the the vr application
25:25
Um, and then obviously the person Involved Then I Opened the vr application with a headset They they they they stepped inside the data
25:41
And the bottom one here shows, you know, they're walking through the cross sections and actually Mostly this is this is done the the scale is it is totally adjustable in real time But most it is done in one-to-one for this type of work. So you're actually working In real scale against the data
26:01
They created a mesh and the way it works is the we the mesh editing. Um The mesh is obviously a triangulated mesh each of the nodes can be moved individually or you can move A subsection of the mesh or you can move the whole mesh Uh, so you can map the mesh very very closely
26:21
To the visual color contours that you're seeing in the data Um, you're moving these these nodes centimeters either way to create the the most optimal fit, uh, and indeed the the result has been proven to be Uh much much more accurate
26:41
against the actual data than the work they could get than the results they can get from From doing this on the 2d screen and automatically creating the the profiles Uh with that I think i'm probably going to have to stop because I think i'm out of time
27:02
Hello Probably help if I unmuted myself Yeah, i'm getting i'm getting coached i've had to get coaching skills through this that was good, um
27:22
That uh Man i'm trying to see if we've got any questions Questions We don't have any questions, but we're getting a lot of interaction in the chat. So um Can it export out the data? Can you export out the data into another? piece of software would need to stay within
27:43
verges, uh It's it's all the data is always kept in the native mode. So for instance, okay If your data is a vector file It reads the vector file And then if you edit it it writes it directly back into the same director vector file same for meshes The rasters are just brought in in native mode
28:01
So actually no it can it can export automatically back straight back into the the file that you got the data from Okay, cool Well, awesome. Well with that We're almost up with the next speaker, uh anything else anything else you want I had
28:22
Uh tie-in no, I think that's Cool Okay Awesome Mr. Harwood, I appreciate it. That was good. Thank you. Thank you for the time and then yeah moderating it all excellent
28:41
Oh don't so let me uh, you turn this off. Yep. I will pop you out of here and there we go. Uh