LiDAR in QGIS, the beginning of a beautiful friendship with Tibor Lieskovsky

all right we are live um good day to the qgs open day community uh thank you so much for joining us um today we're going to have really an amazing session where tibor is going to show us lidar um in qgis and all the amazing things that you can do with it in qgis at the moment so thank you for joining us and we will go on this journey we will be showing some demonstration and then we will have a little bit of a discussion and please ask your questions in the live chat so that i can ask them and thank you for joining us today of you to youtuber uh thank you thank you for the organization and i think that i was thinking how to call this presentation so i decided it will be called by the brief course of eastern europe english or gophenic style english but we will talk about the lighter qgis and things what you can do and it will be kind of improvisation we will talk a lot of about the point clouds then about the lighter processing in the qgis with whitebox tools and vlc so probably i will start with sharing of my screen it seems that we are online perfect okay and so this is the name of the presentation uh first i will say something about the slovak user group uh then uh first half will be about the point cloud thank you please um do you mind just clicking hide on the little streaming notification at the bottom of the screen yes awesome thank you okay and then we will play with the point clouds uh i will try to teach you how to combine the point clouds with gis data for instance we will try to make those kind of visualization of the his combination of historical plans historical aerial photos and point clouds and the second half will be something related with the interpolation of point clouds in uh qgis something about the contours something about its move the smoothing of relief and so so we will see it will be kind of improvisation like i told at the beginning i think the the point clouds in qgis is pretty new thing it was introduced in the beginning beginning of this year and it was one year of development it was uh based on really successful crowdsourcing and it was developed by lutra consulting and a few and other companies and it was developed because of you because of users uh the crowdfu crowdsourcing was really successful and luther consulting like one of the author make also really nice blog about it i think three weeks ago because as usual everything at the beginning has some kind of beginning phase some things are a little bit not working now or not so let's call it not working but for the beginning a little bit complicated but we will go through it uh i will try to follow some instruction from the block of the lutra consulting i ask amy then we will put the those link to the description of the video and for instance uh we will talk about how to install the qr js to make the lighter data working then something how to set up the two two-dimensional or three-dimensional view of qgis then something with the better light either lightning and visualization and so and so and so so we will start with the topic how to make the point clouds working in the qgis uh it's the problem that the points clouds are so new and they require so much new liberal libraries and new things that for now the versions of qgis the point club was introduced in 3.18 are splitted to the two parts one is classical installer without the support of point clouds second is let's call experimental or new installer with support of point clouds so we will go to the web page of qgis i'm working on a really small resolution so for me it is kind of hell but i hope that it will work we will go to the download download page uh those are classical options for more advanced user are the osg of uh installers for the classical user are better the standalone installers uh still i'm talking about 3.18 uh last long term is 3.16 bar about no point clouds there and if you wanna try the point clouds then you have to scroll to the bottom of the page and then you can download the osgo installer or the sound standalone installer modified for the windows to work with the point clouds i will not install the whole rgis but i will maybe show for the people what are not experienced with those installers how to deal with it so you have to download it then you have to start it i really like the advanced install because then you can install the nightly version of gis you can install the different version of the libraries like the gdal pedal and so then you have to check the install from internet and what is really important if you have previous version of qgis installed without the support of point clouds if you have for instance the version 3.16 or something like that then please change the default part to installation because what can happen that it will be installed in the same folder and the libraries can be mixed and those should be really problems with the files so my advice is if you have two installation give the new name something like that then you can skip those things and you will see the list of resources you will choose one and then you will see this kind of table and if you want to install just qgis then go to the desktop if you wanna install the latest or current version then go to the qgis desktop and click here and you will see the latest version is 3.18.2 instance i really love to play with the developer versions with the night nightlys because every time our new feature there so it's nice to have it but sometimes new mean sometimes problematic so it's up to you and uh you can see that you can also install the saga or grass keys here but if you will choose for the qgis it will automatically download the dependencies via this installer so then you will just click next next next and you will wait a few minutes or hours it depends on the speed of your connection and everything will be okay so this was about the installation so please use the uh not the installers from the front or from the top of the page but from the bottom of the page and i believe that soon we will have again only one version of qgis okay so i have the qgis but another really nice software what we will need is the cloud compare and why uh i will try to explain you can find the lidar data in many forms for instance i will provide a link where you can download the data from the slovakia those data are in really nice accuracy it was estimated like under the five centimeters in the horizontal and vertical precision really nice density like 20 25 points per square meter and points of the ground after classification and what is really nice that those data are really free uh if i will go to the our national geoportal then i will go to the map client i will choose the terrain then i will go here and here you will see the data export and i can mark the shape you are limited with the two square kilometers those data are really free but this is kind of technical limit you can make the shape here if you are under the two kilometers two square kilometers everything okay but i forgot to tell you you have to switch the point cloud and then you will put the email and then you will check that you are agreeing with the terms of use in a case of slovakia those are only two rules that you have to provide the source of the data in the description and you you cannot share the original data what you already downloaded have several practical reasons but it's up to you many of countries have their own national authorities with the freelander for instance there is the really nice database of the lidar of united states and those are indexed in the tiles so we will see but if somebody want to play and it is from country without the data we can download the data from slovakia uh then we will receive the mail with the link to the data and there will be zip file and in this zip file that's the one of the first problem that i have to solve in the pragmatical point of view we did how to deal with lighter data in qgis if you will download the zip and point cloud there will be not one single file from the area of interest but there will be few of the stripes of the lidar data it is called lazid like zipped version and what is little bit problematic that at the beginning you can deal only with one file uh with uh with the one file of lidar data in gis you cannot work with the merged uh onfly merge loss files so for this first problem we will use the one open source really nice open source software and those this software is called uh cloud compare i can solve the problem how to merge those data into one file by many many ways for instance really nice tools are lost tools but they are not completely free and they are mainly for windows then another alternative is to use the pdl and the workflows and pipelines but there are the pdl are really nice and powerful but maybe they are for a little bit more advanced user so we will download the cloud compare it is from the first link then you will go to the download section and it is not necessary to have the latest version it will be nice to have the stable version if you are on the windows then you will go for the uh this installer then you will download this software uh we will install this software and then you will have the cloud compare somewhere in the computer uh you can type cloud compare or i have it here and how to work or how to merge the data you will take the data uh everything is finishing with l a z or l i s what is a uncompressed version of later the glider data you will do drag and drop to the mainframe it's one of the op way how to do it then it will ask you what do you want to import from those point clouds like the time number of returns and so and so uh usually i'm just hitting the apply all then what is interesting it is asking you if you wanna shift those data the reason why you want to shift those data is that usually the coordinates in the projected systems are really big you have several number be between the comma several numbers behind the comma and it is a bit complicated for the memory of the computer and handle it so it will advise you to make the shift to the local coordinate system and it is really safe because you can do everything in the local coordinate system and if you will re-export the results it will automatically put back those offsets so my advice is just hit the yes and yes all okay we will wait a little bit i hope that not so much and first thing what we will do is that we will merge those data it is really easy you will hit the shift and you will select everything everything means also this folder and it will make the protest against it but it is not so important and then you will go to the edit and then you have the option the merge and no you don't want to calculate the original point cloud index and if you will hit yes okay i already calculated but it was necessary you have the one point cloud here a little bit merged and you can play with it first thing what you can do is save those merge files you have to select the file then go to the file save and now what is really important you have several options how to save it usually the most common is loss or lazard and here is one dangerous option option if you want to play with the qgis because uh that was also usually my mistake that here is the option to save it in the newest format but they have some bug there and it is problematic in the qgis so please don't select this last point one three or one four uh select just last cloud and i will go to my default folder for this and i will call it a pc like point cloud and merged it will ask you about the resolution usually i'm hitting the okay so we will play with around 10 millions of points uh cloud compare is really powerful tool you can make really nice things you can make the really nice visualization like for instance you can visualize those data by classification of a divider you can change the colors for something i don't know something more something better you can play with the things like the profiles cross-section and so and so but i don't want to talk much about the cloud compare because it is related or we want to talk about the qgis but for instance those are the nice video tutorials of the how to deal with the point clouds in the cloud compare i really like the channel of android i will introduce this link he made the video how to work with the cloud compare and somehow with the qgis okay so back to the cloud compare because what should be really interesting in the also in for the several reasons is that as you maybe know if you have the so-called classified point cloud classified mean that every point in the point cloud have the class like it is the point from the ground it is the point from the building it is the point from the high uh vegetation low vegetation and so and so and also you have a lot of noise in the lighter for instance this is underground noise and head and it is something what we don't want to have and usually i'm using the cloud compare also for splitting those points by the categories so for instance i will i want to separate the buildings and ground from this point cloud and then we will separate the high vegetation point cloud and we will save them separately uh i will try to show you why okay so we have a merge point cloud let's separate it uh then we have to go to the okay i don't know where but usually you can use this icon filter points by value you will see later the what are the numbers of the classes there are something like international standards so for instance i know that ground is a point class called two and because i wanna select just the classes from number two to number two i will give the number two and then i will hit the split i will wait a little bit and now what we have is the another two files one is everything without the ground point cloud second is just the ground and i will save it like the file save and i will save it like the point cloud and it will be not merged but ground and i will provide extension la is okay uh then i will for instance separate the buildings i will clean clean this up okay i will go to the buildings buildings are class number six split okay not this one this one those ones are buildings i will save them and then we will do last things and it is the high elevation okay so last one the high level vegetation i think that that is class number five split okay here we have our trees branches and so later it will depend on the time we will we will play with the interpolation of those data and we will try to do this uh interpolation in the whiteboard tool whitebox tools in the qgis but what is good to know that you can make the interpolation also in the point cloud cloud compare but i prefer the gis way to make of the interpolation okay okay so let's go back to the qgis okay we are here and for instance first things then you can see if you have correct version of qgis for the lighter is that when you will go through the browser to the folder with your data if you are see the point clouds here then you are okay so at the beginning we will add the all point clouds the merged point clouds and what is important uh the point clouds are in the lis format about the qgis use something like internal format internal indexed format it is also one of the standards or it will be one of the standards and it have to rain decks and convert those files to that format it will take while ah okay i have some problems okay i will try to another directory okay there was some problem i believe that it was related with the uh extension of the format uh it's good to know that maybe the las files are problematic uh we will try to check this issue and maybe if it is really issue of the queue qgis video reported if it is the issue of the the cloud compare we will report it to the cloud compare but i have the same data but in the slovak version the merged mean resolution because i was showing this to my students a few days ago so if you will first time but i will for a moment i want to simulate something and i will explain this little break later okay so i have here the from different exercise but we will rename to the listed merged this the account this to the buildings it was made made by the same way in the cloud compare but with the different extension called la s li that and now what is important in remark you cannot see another folder here you can see just the file you will add the data to the by the browser or by drag and drop to the qgis we will use the merged and then you will see that it is start indexing those data and then it will start to convert those data to the another format and really powerful format for the displaying of huge amount of data and you will see that from that moment we have the new folder here it is called the same by the same way how was the original file called but they have the prefix apt and those are here are the indexed and stored data so for the later if you for instance work with your colleagues and if you want to share those data it is necessary to share not just the original original the last file but also the the folder related with those files okay and first we will inspect those data in the two-dimensional view uh this way how to play with the styling of the data is that if you will hit hit f7 uh okay here you have the panel of lifestyling i have a little problem with the small screen because lower resolution but probably for the people who are working in the gis they know this panel pretty well and what is important to say that here we have several options how we can display those data first icon is for the two-dimensional view uh you can switch off the classes what you don't wanna have uh what i really miss is that something like you will click with the right click and you will have the option the deselect all so i will do it manually i wanna keep just the grounds i will turn off the vegetation the low point noise reserved water and everything okay then what is really interesting or really good to know is that uh here if you will scroll oh it will be in the 3d mode i believe so you can for instance switch the the shape of the point cloud from the squares to circles and so and so maybe later i will show you the trick how to make the visual of visualization of this data a little bit more interesting because you can see the that here we have some buildings here we have some trees but it's really hard to distinguish the shape of the rooftop or the shape of the tree but we want to really play with the 3d view uh so if i will go to the 3d view then i can start it with the view and new 3d map view but we have to set up a lot of things at the beginning what we will set up is the how this point cloud will look in the 3d and we will try to solve the problem that every time when you will start the 3d view you have your nice 3d point cloud but unfortunately you have the two-dimensional plate under it so let's start to solve this problem i again hit the f7 and first i will change the appearance of the three-dimensional point cloud as you see even we switch off or turn off the classes like the noise low vegetation and so we can still see it and the reason is that we have the second tab for it this is for two dimensional displaying and here is for three dimensional and again we have to uncheck everything what we don't wanna use i will keep buildings i will remove the noise and so and so and so okay so now you can see that you can see just the trees the ground and the buildings but it's one problem here that you can still still see the two dimensional data under the point cloud so you have to go back to the two dimensional displaying and you will turn off the displaying from classification to the extent only so we will go here uh then you have just the symbol of the bounding box under the point cloud and if you wanna turn off the symbol then uh go to this sometimes i don't like the behavior of the panels in the in qgis now i really want to know where is my panel okay quick and dirty solution it's the same like you will hit the f7 it is the properties of the data i will go to the symbology and i will for instance turn off the simple line or i will put the open and so okay it's much time better and maybe you want to display those data in the more interesting way and for this we have the two options and those are the show shadows and we have the another option show item lightning item lightning is the something like ambient occlusion but for the point clouds ambient occlusion is very how to display the three-dimensional objects with the and with the highlights and the more exports things and so what is a little bit this advent teach of this implementation but i know that it will be fixed in one day that if you will turn off the shadows and item lightning then you will see something like displayed under the point cloud till now there is no way how to fix it but i believe that there are some ways or it will be fixed later and about moving about in the free 3d viewer you can use the mouse like scroll wheel for zooming through out if you will click on the scroll wheel then you can rotate if you will use the left button you can make the pan and so and so but sometimes i really like this compass and this old-fashioned way how to rotate the data because sometimes there are more precise also in the block of lutra consulting is written one interesting way and it was inspired by the old computer games and i believe that blender have the mode how the same mode how to walk the through the 3d data so what we know now we know now okay it was more chinese than english how to display the data in the 3d view first we have to what we have to know is that you have the separate tab for this you have to turn off the classes you have to turn off the two-dimensional displaying and what is interesting to play with it is that you can play in the 3d settings with the point size and maybe if you have a really big amount of the data you can try to play with the this point budget this point size is pretty simple it uh deal how to display the size of the point in 3d and now this point budget mean that at the moment is displayed one million points uh on the screen uh it is kind of the generalization if you wanna have the more points maybe for the smoother surface or something like that you can go more for instance the two millions but then you have to be careful with the power of your computer and the speed and other things like that okay uh let me combine the those data with the raster data i'm thinking that if i will stay in this project or i will do the same for the another project because i promise you how to make the the combination of gis data historical maps and so so so so maybe i will no we i'm talking too much about it so we will not combine it with the historical data and with the historical maps and historical rl photographs because the workflow is the pretty same uh what is the advantages of those data or or of the viewing of 3d data in the qgis if you will compare it for instance with the cloud compare is that in the cloud compare you can view only the point clouds 3d object but you cannot view the gis data so i will try to add the digital elevation model here so i will go to the browser i will add this dem by doubleclick i have my two dimensional display of the data here and we can make this model or this raster a little bit better the the raster was derived from the points cloud and it was derived just from the class called ground maybe we will show it later how to to display it how to interpolate it so first thing what i want to do is that i wanna check the coordinate system it is something really related with the slovak issue so i don't want to explain it i will set those data that they are in the coordinate system and one way how we can display the data is that we will use the hip symmetry that mean that we will display it by the colors not multiband color but single bond play with the color i will choose one of those color arms or i have maybe something better okay but still you cannot see the details the morphology of the relief so good way how to display it is to make the copy of this layer you will go to the duplicate then i will rename the this the dem will shade and again i will try to press f7 but without the success so i will go to the properties of the dm hill shade and i will use the let's let's call it on fly generate hillshade uh i will for instance add the more dramatic way i will i hate those kind of words for instance for me it's bloody hell the difference between spatial and spatial so okay so we will set this factor and we will turn on the hill shading and here is the one of not the problem but you can see that because the layer is at the top you can see the hill shade on the top and the traditional approach how to combine the hypsometry with the heel shade was the to setting the opacity like for instance for the 50 percent but this is not so good way because then it will wash out the colors and those kind of things so much time better way is to set the opacity to the 100 percent then you will go back to the styling and then you will set the blending mode and you will use the multiply and it seems much time better but now we have the two dimensional hills model of this castle and we will now put it to the 3d here so then you will go to the settings you will go to the configure and you will take the first option that you don't wanna display the flood earth but you wanna go to the dem or dm uh this error mean that we didn't choose the digital elevation model so it will be draped on our digital elevation model and we will see what will happen okay so now it seems that our model is in the 3d uh we have some problem with the outliers because it was cropped but you can set the no no data to those triangles and there are a lot of things or a lot of way how to solve this and maybe now it will be nicer to turn off the class the ground from the 3d point cloud so i will go to the merged what does it mean the point clouds i will go to the symbology i will go to the 3d view and i will turn off the two-dimensional ground but the i will turn off the class called ground and now you can see the data but if you can as you can see the data have a little bit of generalization there and also we have some problem with the offsets uh maybe we will try to play with it or we we will see i want to show you how to make the the underground data or underlight data a little bit sharper it is good that there is the generalization because imaging you have now the small screen and also you are watching in the small screen and you don't want to have the every pixel in the display so it is good but sometimes it is better to put it to the regular way so if you want to increase the resolution of the dem in the 3d view then you have to go to the options you will go to the configure and you will increase the tile resolution i will try maybe 256 and the digital elevation model is now much time more detailed and also uh i want to increase the resolution of the displayed pictures like in this case it is the hill shading but it should be also the map or it should be the url photography maybe we can try to play with it for the one second here we have some data from slovakia we will wait for the refreshing of this 3d view so if you want to increase the resolution of the raster data then you have to go to the configuration you will go to the advanced and then you have to downscale this max ground error it is a little bit contrary into it to you it is really hard to understand what mean what so for instance the resolution of the texture will be something like point three and as you can see it is much time more detailed i'm a little bit surprised that i put the rl photography here and i cannot see the arrow photography here but after this video of the digital elevation model yes i can see also the the rl photography here uh you can turn off the classes for instance you can inspect just the uh if you can inspect just for instance that you wanna have just the buildings there you will switch it from the point cloud from the symbology from the 3d view to the just video here just building is here and so and so and so okay uh it is a lot uh those are a lot of things what to talk about the displaying of 3d data for instance you can make the really nice animation here you can play with the elevation of sets and so and so but i will go for next i mean do we have something like 20 minutes yes no if you want to continue um please do if the stream does stop i will just restart it and um if the viewers are happy with that if they can throw me a comment in the live chat then we can just continue i'm finding this fascinating it's brilliant so please continue okay thanks so i think that it was somehow enough of displaying the 3d data and 3d view maybe we will play a little bit with the combination of the two-dimensional view and the gis data why the two-dimensional because i really love the 3d in the gis but i'm sometimes it is just for the effect every time i'm trying to explain to my students that if you have nothing to say then give the 3d picture and if you have absolutely nothing to say give the 3d animation unfortunately it still work but what is the biggest value of those data are the two dimensional representation so we will go back to the i will rename this to the point cloud and for instance we will try to compare the dose data with our slovak cadastral or with the openstreetmap we will see so i will go to the properties f7 is much and easier i will go to the classification unfortunately i have again checked out everything and we will keep just the building and high vegetation okay and i will put it to the top okay and for instance i will try to it because i wasn't working with this for a long time the the vector tiles of openstreetmap i really like this plug-in uh map tiler so i will try to add the street we will see by double click ok map tyler yeah we have the streets but we don't we cannot see the buildings oh it's a little bit complicated i will give there the slovak duster but you can play with the open street map data okay here we have our cholesterol map and you can see the shifts in the data it is related with the coordinate system with the generalization of those data with the things like that we have the cadastral maps but from the beginning of the 20th century and we have the remedy and so and so and so but you can compare the footprints of the buildings you can use it for the digitalization of the open street maps and many application so maybe i will show you some more tricks how to display the buildings in the little bit better way to understand what is happening there but unfortunately till now we have to use a little bit of work around but i hope that it will be solved in the next versions so for instance i want to add some texture to those buildings to the rooftops of the buildings and if i will go to the properties of the point cloud uh then in the two-dimensional i can see that here i have the classification but you can also the display by you can display those data by another ways like the attributes like the intensity returns number and so and so and what i really like to combine with the displaying of the two-dimensional point clouds is to play with the intensity of the data uh so i will add the just the buildings from my point clouds that was the reason why at the beginning i was splitting the the files to the buildings and the high vegetation and so and so so i have the merch vegetation okay i have buildings here it was generate really quickly i have just the buildings and maybe i wasn't playing with this for a long time so maybe i will be not successful then i will finish i will go to the symbology and i will not take the classification i will take the attribute by ramp and you have here something what is called intensity and if you will choose the good color ramp i will play with this for instance you can see okay it wasn't so good color ramp you can see that you can play with the things like the displaying of data by intensity but you have to play with the intervals of the how to divide the color scale symbology and so and so but unfortunately i didn't prepare this for myself so maybe i will talk about it in the later speech but it is one of the way how to display those data in the different way maybe it should be useful for instance for the i have the vegetation here yes i have now you can see the trees like just like the just the with the same color with the same uh two-dimensional color so also we will add some texture we will use the same trick we will go to the symbology then we will go to the attributes by ramp also we will take the intensity but i will try to choose the different ramp for instance this green and now it is not so good color ramp but if you will played it with it if it will be not just the black and white run but maybe something complicated uh or you will play with those intervals you will add some more for instance i will try to add some black color here you can urge you can have the different results and different way how to view those points in the in the two-dimensional display for now it is bit complicated and it was the it is one of the reasons why i didn't prepare this really well because for now you have to split the points the clouds to the buildings to the low vegetation high vegetation and you have to make the style of each layer or each class separately but as i know from the developers of these point clouds they wanna have the possibilities like i hope that in the next versions you will have the options like the make the filtering of the points by the point clouds by the point class by the for instance elevation by the intensity there's something really similar like we we are querying the classical vector layers and there we have select by attributes and so and so and maybe i promise to you that i will play also with the way how to interpolate data to two dimension to to raster but i thinking that now it is enough of information for for now uh maybe if people are interested how to interpolate the digital elevation model in different ways with the different options like linear interpolation clicking id how to smooth the surface how to fix the imperfection of the surface i will prepare another topic about it for the next would be absolutely amazing please do we would love to have this kind of you know interactive um session for the next open day that would be amazing okay so now i think that i will finish and it is time for your questions if you have some all right there are actually quite a lot of questions um i would like to ask the viewers if they have a specific question just to post it in the live chat now so i can ask but i will go through the live chat and just there have been some questions along the way um so first of all there are quite a lot of people saying ahoy from slovakia so that's pretty cool and then and then um so there was quite a bit of talk um so this video will be recorded and i will put it up um obviously on the qgis channel and um tibor will send me all the links to the data you need um to the webpages you need and i will add that into the description so that when you watch the recording of this and you want to sort of do this yourself in a practical manner you have all those links at your fingertips all right so i don't want anyone to sort of worry about that there were quite a few questions about that now we're going to go into the more um sort of technical side of things there were quite a few questions while you were um sort of manipulating the data into the different file types so one of the first ones is what are the advantages of ept over a lazy the file types i think the this is the question for the developers they have big research how to make the the displaying of those data really fast because what do you have to realize that we are talking this was the small amount of data this was just 10 millions of points but usually dealing you are dealing with the hundreds of millions of the points and it is related with the indexing of the data with the quasidratric tree structure and so and so so it was the decision of the developers to use this apt format and for instance probably the apt format is really good because like i told at the beginning the us have the later lighter data for free in the cloud because the amazon offer for for the projects which are offered the data for free they are offering the free data storage and you can directly download the those format in apt from the cloud and i hope that one day the slovak authorities will also move those data from las format to indexed ept format in some way of the cloud because then it is not so necessary to download the data i have to work with i'm working with slovak data and amounts of data are in the terabytes all right and i i do want to just thank um luigi in the comments for answering and explaining to those different file types um and there was a question on where this is so what sort of fort or castle that is do you know uh it is somewhere in the slovakia i think that it is in the south of slovakia or in the eastern of slovakia i'm working only with cultural heritage i'm working with the archaeologist for 20 years what i'm making for the archaeologist is this kind of special visualization of the lighter because for instance i call it the archaeological archaeology of invisible now you cannot see nothing here even if you are if you have the hill shade but if you will turn the turn on the special visualization uh what i'm making with a lighter with the raster representation of lighter what you can see is something like stone hitch but it is bigger than stonehenge and it is two 2000 years older so all my applications are related with the cultural heritage with the castles with the things like the hill forts and so and so that's brilliant you're gonna have to be giving us an episode of time team next all right so um there is a question uh what is the preferred hardware configuration for working with point cloud data uh i was a little bit surprised that it was not necessary to have something like bloody powerful pc uh uh i think that the enough of ram uh i was surprised what you you can solve with 16 gigabyte of the ram i'm using the 64 gigabytes of the drum and those are only rare things when i hit the limit and my advice is especially if you are working with the rasters really really big rosters play with the ssd drive and try to go to the optimalization like to have the tile the raster and to have overviews in the raster but i was nicely surprised like i think that little bit better common computers are okay for it ah fantastic all right so um uh i think it's s1 keller it's a slight more technical qgis question so they say it seems like the 3d surface has a certain height over the ground and and the center in quotation marks of rotation is at the 2d ground is it possible to move the center quotation marks of rotation up into the 3d surface i understand this question but i believe maybe if it will be martin here the developer of the 3d viewer he can reply the those questions but i think that they put the google summary of code they put the two proposals one is for the on-fly raster visualization and analysis and second was related with those 3d options in the 3d viewer and i think that one point was to make to have the option to shift the point to shift the point of the rotation so i believe that for now no but i i believe that soon we will have this option okay i mean this is kind of the cutting edge brand new version of tgis so it's amazing um you know what you can do and i mean obviously things will grow as this is used more no yes it's really necessary to to remark this because i think that developers of qgis make really nice piece of work that imagine that they will put those complicated things like the point clouds within the less than one year to the software and it is just the beginning because as i know they wanna later add the options not just the view the point clouds and lighter data but they want to add the options to edit those data crop those data to make some analysis and but you have to work step by step and you can have the options for editing the data and analyzing some data but if you have an option to visualize visual to see the data it is not useful so at the beginning they start with this and they will continue with more and more functionality all right um so there's a question here from shane carey um are they asked if it's possible to measure um vertically or say get a building height in the say 3d viewer for example so can you measure in the 3d viewer the vertical height of an object uh i'm not sure because i'm using the 3d viewer just for the 3d free review and previously but for sure you can do it in the cloud compare okay okay it does seem that you are able to measure if i read a little bit further down in the comments um shane does say that he's able to measure in that view but um the vertical height didn't seem to be correct i'm assuming if you get like four kilometers for a building it's not right so i think there is the option but um it's gonna see what like the accuracy is or something along those lines mm-hmm all right um oh here's a great question uh just from someone called am is it planned to allow drawing of the vector um or relaying of over a point cloud or more interaction between vector and point cloud and raster data so is there any plans for this as i know from the developers those are the plans because they are receiving the request for all kind of a way to make the 3d vectorization of the buildings in point clouds and so and so but like i told step by step um so andreas uh says that he'd be interested in the interpolation method summary so i'm assuming assuming we could address that in our next session um the interpolation method summary so that would be a great sort of place to go um luigi asks did you find a size limitations uh yes for instance uh in the 3d viewer it is the limitation even with the underlying rosters or two to the dimensional things and i cannot clarify this by the exact way but yes for the uh bigger area i have the rasters with the lower resolution like because usually i'm working with 25 centimeter per pixel of the diet resolution and this is a lot of pixels so if i want to display something like this this is a few kilometer square kilometers then i'm using the lower resolution of the data and i think that i didn't find the limitation with the point cloud but this is the advantages of the apt format and the approach what they use because like i was trying to show you that in uh properties and in this 3d view here is the point budget and that means that you can have 100 millions of the points but still on the display will be something like 2 million points and i think that i reached the limit in the somewhere in the 5 million or something like that all right okay brilliant um then the um s1 keller asks just to confirm is it possible to load ept directly into qgis uh good question and i don't know for next time yes for next time all right and then um the last question i'm going to ask it's um sort of the end of the live chat and we're also sort of pushing on an hour and a half so um the last question is again from am um is there a plan to add manual classification tools for the 3d view uh it is also the question for the developers unfortunately martin cannot be here but i believe that yes and i also wanna i i also really wanna have this option because for instance i see the problems like in this very how to displaying the castle because the usually castles are ruins and ruins means eroded walls and so and the classification algorithm remote all points from the with the ruins from the point cloud because they cannot understand that those are the buildings but eroded buildings so that's the reason why also i need the manual classification interview so one day i hope but maybe also what is really important to remark that even we are using the open source the the amount of the time spending in the development is huge those are the hundreds and thousands of hours so it is not nice to put the developer just a feature request but please if it is possible if for instance you have some kinds of companies that are leaving from the open source please put some money to those kind of campaigns or crowdsourcing for this or if you are student or if you are out of budget try to help the people with those kind of presentations with the answering of the under the forums so please don't just ask the feature but try to do something for it fantastic um i just want to thank luigi for answering that question in the comments he says you can import the ept but you have to use a plug-in to do so so it's kind of outside qgis but it is you can do it using a plug-in apparently he has mentioned the name [Music] um could also be zobruff castle apparently it is possible no um thank you so much tabor this was fantastically interesting i would absolutely love it if you could come back on the next open day and give another presentation on this you know the 3d um being able to picture stuff in 3d is so user-friendly and so you know it gives such a great you know when you presenting a map to someone people are impressed when you can see everything in 3d and it gives a great idea of what's going on so thank you for doing that and thank you for the opportunity absolutely and i'll be in touch um just for everyone who's worried i will as i say i said before put the links into the description um as soon as tabor sends them to me and um hopefully we will be seeing you again soon hopefully so have a nice day or night every everywhere where you are stay safe and i hope that see you soon thank you bye all right i ended the stream so

Channel: QGIS

Views: 2,371

Keywords:

Id: SsoRDfhZzm0

Length: 75min 0sec (4500 seconds)

Published: Fri Apr 30 2021