Automated Linework from a Point Cloud Using Linear Feature Extraction

JOHN SAYRE: Hello, my name is John Sayre, technical marketing manager civil infrastructure here at Autodesk. Today I would like to present CES601587 Automated Linework from a Point Cloud using Linear Feature Extraction. I will start off by thanking you for participating in this session, and I hope you find it informative. Before we get started, though, in this session, please take some time to read the safe harbor statement.

All right, so the agenda for today, will go as follows. We will go through a quick history and timeline of the ReCap Pro improvements from 2023 to present. We will only go back to 2023 because that's when the ability to actually publish and view a point cloud in the cloud was introduced.

I will go back and do a quick demonstration of how to get point clouds to the cloud using the Publish command inside of ReCap. And then we will go through a brief walkthrough of the cloud user interface to make sure everybody is familiar with it. We're going to go back and do all of this just so everybody is starting off on the same playing field and that we can all start from the same point before we start to go through the actual linear feature extraction. There's been a lot go on in the last couple of years. And I think you'll see that in the history recap.

Then we'll get into the meat of our presentation and step you through a couple of different workflows for pulling linear features from a point cloud in the cloud all right. And lastly, we'll do a quick summary just to wrap things up and just talk through what we've already gone through and what the expectations were. So we got a lot to go through. So let's jump into the history and timeline of some of the ReCap Pro improvements starting in 2023 moving forward.

All right, so starting in ReCap Pro 2023, users were able to publish structured or unstructured ReCap projects to Autodesk Docs, Autodesk BIM 360 Docs, and Autodesk Drive. You were able to view and mark up point clouds and view the 360 real views of these projects directly inside of a browser, all right.

Just to be clear, you could view point clouds that were collected with a scanner or point clouds that were collected by a drone, all right. And you can see those in the ReCap browser. We will take a quick look at the interface of the browser. We'll do that in live here in just a minute.

One thing I'd like to point out, you'll need to use the desktop connector because we're going to be using the Autodesk Construction Cloud, all right, so Autodesk Docs more formally. You do need to have the desktop connector installed in order to get the publishing process to work with ReCap Pro.

All right, so the desktop connector that I'm using is 16.4.0.2062. Now, if you have a version of desktop connector that's 16 dot whatever, 16.0 or something between the version I have, you're good to go. It will perform great, and you'll be able to publish your point clouds to the cloud.

If you have a version that is a 15.8, if it's anything 15.8 anything 15, you need to move forward to the 16 dot version of desktop connector. So the bottom line is, is that you can publish a point cloud to the cloud, which is a much more efficient way to communicate and share point cloud information. All right, we'll move on here.

So in ReCap Pro 2023.1, they gave users the ability to automatically classify scan data into ground and non-ground points. You can select one of the predefined level of detail options for processing the data for terrain and features, or you can select Custom for additional options.

ReCap Pro 2023.1 also displayed the point cloud in different colors based upon that classification. You can see that here in the example. It makes it easy to distinguish the ground points from the points that are not classified. So the ground points, here in this image, are a brownish orange. And the ones that are not classified, are green.

And those are the trees and things up along the highway You've got some greys that are there, which would be the cars, things like that. They're not ground points. They're not what we're going to build our surface from.

So having the ability to filter the points by classification, it allows you to export only the points you need to build a surface. So it shrinks the point cloud down so that you're not hogging resources whenever you have that cloud-- open it on your machine, all right.

Using the classified point cloud to generate a surface inside of Civil 3D, will ensure that you're only using the points that you need to accurately represent the existing conditions or the ground, in this particular case, for the surface. So in ReCap 2024, there was more development that was extended-- it extended the classification management and automatic ground classification to scans imported from different types of raw scans.

So we're able to actually use more different-- or actually use different types of raw scans from different vendors, things of that nature. So that was one of the things, in 2024. And the main-- well, I say the main thing. The main thing that we're going to look at today was linear feature extraction. You're able to actually get into the ReCap viewer, and you're able to actually extract geometries of linear features with the help of machine learning and a heuristic algorithm, all right.

Bottom line here is that you're able to quickly generate lightweight surface models, meaning, in this particular instance, you can see in this image, we're pulling this wall from the cloud itself. So in the end, we're only going to bring the line work in and generate an existing ground surface from the line work, not this huge point cloud, only the line work. And it's just as accurate and very, very lightweight.

Now, what we're going to do next, is actually get in and show you-- I'm going to show you live what I'm talking about. So let's get on with the show, and let's take the point clouds to the cloud. And we're going to do some linear feature extraction, all right.

First off, though, what we're going to do, is we're going to walk through how to publish a point cloud to the cloud. So let me get into ReCap here. And I have ReCap Pro open. Now this is ReCap Pro 2024, all right. And what we're going to do-- I've just opened up the point cloud on my local machine, all right.

In this version, all the way back to 2023 when we were able to start publishing point clouds, you simply go to the Home tab in the ribbon and move to the right. And this button right here is your Publish button. Now mine says 20 days. Ago that's true. I did publish this particular point cloud 20 days ago, all right.

But we're going to act as if it's not been published. And I'll show you the difference here. All right, so it's going to initialize. Notice it says Update or New. So this is going to allow us to actually push an update if we have published it already. So we could have added additional scans to this particular model inside of ReCap Pro, and now we're going to update the model in the cloud.

We would then hit Update, and all you simply have to do is just select Update, and it will start that updating process. Now, if we select New, then we're going to be able to navigate out and place it in a new location. It'll be a new publish of this point cloud.

So we're going to call this road rehab, and we'll call it-- how about this. How about we call it AU road rehab. We're going to go out and select our Autodesk connected drive location first. So I'm going to select that. And I'm going to navigate out to-- and let me scroll down here and show you --Autodesk Docs, my infrastructure technical marketing, my ReCap Pro Cloud Projects, and then AU 2023. And I'm just going to select that particular folder, OK.

Or I could select a new folder, and put it in there. So I could say, hey, this is for my AU presentation. I'll select my folder. Does not exist, obviously. So I've got to create a new one. Let's backspace this out. Hit New folder, and call it AU presentation, which is fine, all right.

And I'll go ahead and select that folder. Now, if this happens, and it's OK if this happens, it says select a valid Autodesk connected drive. Because the desktop connector has to build that folder, sometimes you'll have to back out a little bit, come back into AU presentation, which is the folder we want, and hit Select Folder, and everything will clear up. So it does take it a second to push that folder for the desktop connector.

Now, in this particular instance, we only have one scan. We can have this one scan or we could have 50 scans. It doesn't matter. This particular instance, we only have one, but you're able to check, on and off, any of the scans that you want. So if you only want to push to the cloud, maybe, five of 50 scans, you just select those five.

But in this case, we only have the one scan. So we're good to go. We can simply hit Publish, and it starts the publishing process. Now, it does take maybe 30 seconds or so for this to start the publishing process. And then you'll get a balloon that says that it is publishing, and that you can actually start to work with the point cloud here in ReCap Pro if you want.

Or you can actually close ReCap Pro, once this comes up. It's getting ready to do it. It says it can take some time to upload to the Autodesk connected drive. What it's talking about here is using the desktop connector to do this. Now, I want you to remember, we are live here in this recording. I am pushing this point cloud. And I am going to let it push through the desktop connector as we continue to do this presentation.

OK, so it's doing it in the background. I'll minimize here just real quick and show you this is my desktop connector, and this is what it looks like. It's pushing this particular RCS file and all of this information while we're doing the presentation, all right.

Now, if you open back up ReCap Pro, you can see that it was published successfully. And I can close this, and I can close ReCap Pro if I'd like. And we'll start off by looking at what it does push up there after it's done publishing.

Now, it is working right now with the desktop connector. So it's not going to have anything necessarily completed inside of that folder and in the Construction Cloud, as of yet. But what we can do, is look at a project we've already published. And this road project is the one that we've already published.

So I can select road project, and then road rehab, this is what you will see whenever you publish. You will definitely have that same support file and folder that you have whenever you just create a point cloud on your desktop. That information will be right here. You will have this PNG file, and this JPEG file. Those files are used inside of the browser to show you where you're at using the project browser. And then you have your .RCP file.

If you want to view a point cloud, you just simply select the RCP file and then we can view it. Does take just a second. All right, so here's our point cloud. Now, we're going to be able to move, manipulate, do things much like we do inside of the browser version, or I'm sorry inside of the desktop version. We'll be able to do it inside of the browser version here.

Now, real quick, what I'd like to do before we start pulling linear features and talking about the functionality, is I'd like to show you just a few things inside of the browser to make you feel a little bit more comfortable.

Over here on the left, is the project browser. This is where you can actually see all of the scans that have been published. That's right. It does not combine those scans when you publish them. If there's 50 different RCS files that are incorporated into this project, that's how many scans that you will see here. If there's one, like what we've got, you're only going to see the one, all right.

If you've got annotations, if you remember, in previous versions of ReCap Pro, you could add annotations. I can add annotations here, and I'll just go ahead and show you right here, with Create a Note. So I can Create a Note right here that says I need-- we'll enter a title. We'll just call it test. We'll say, I need LFE for the curb.

All right, so you're telling somebody that has published this from their desktop to the cloud, you're telling them here in a note, that you need them to pull the linear features for this curb right here. All right, so you hit Save. It saves it in the cloud.

When that person who published this on their local machine opens that particular point cloud, it will sync with this publish. And it will see those notes, and they will see that they need to, maybe, do some things. So they can click on Test here. Actually they can click on Test here, and it zooms to it. And then they can see the note and know that they need to pull the linear features.

So they can see those annotations. Now that works vice versa. If the person that has this on the local machine that published to this particular location, if they have some notes and annotations and things of that nature they'd like to add, you will get them. Whoever opens them in the browser will see those annotations also. You can also delete them very, very quickly.

If you want to see the linear features, after we have extracted them out of here, you can see all of those right here. And you'll see that populate once we start to extract linear features.

Now, down here in the bottom in our toolbar, we have just a few of the commands that we have inside of the desktop version. You can toggle the project panel on and off. We can actually select to toggle the mirror balls on and off. If you have multiple scans, the mirror balls are where the scanning locations were. So you can see those. If you want to zoom to them look at real views.

You can also change the point display, much like we do inside of the desktops. So we can make them real big, make them real small. Three is a good number to leave them at. Now, I want to point one thing out because I do this all the time. I select point display size, and this stays up and then I go to something else. In order to get this to go away, you just select it again, and it goes away, Just a little tidbit there.

Now, we can also control the visualization. So select a different visualization mode. So if I want to change it to just this normal mapping mode, I can see it in a different color scheme. Same thing with intensity, I can actually go in and change to some predefined things like grayscale, so on, and so forth.

I've actually got a selected shading mode. I just use a gray scale, just a shaded mode. And then my RGB color mode, which is what we're going to use today in this presentation because it gives you the actual look and feel of what's really there. But you can change this. And later on, when we get into a different scan, we will change it so that we can see things a little bit better in that particular scan.

Again, here's where you can create a note. We can create a distance measurement. Those distance measurements also go through and are pushed back to the original location on somebody's desktop that published this. If there's a measurement that's shown in the desktop on this cloud, it will also transfer to the cloud version also or in the viewer.

Now, I could go through and give you probably a 10-minute more presentation on how to work this, but I want to get to what we really are wanting to talk about today. And that is linear feature extraction.

What we're going to do, is we're going to extract this curb right here. Let me get back to this curb ramp. Now, this particular scan was a scan that was set up with a backpack, I believe. May have actually been a car-mounted scanner. But as you can see, it does have very, very good definition. And that does definitely benefit us.

So we're going to start by selecting linear feature extraction. And if you're like me, it didn't really do anything whenever it started, but it changed this menu down here. The first thing we need to do is tell it where we want to extract.

So we're going to select right here. You can see this arrow. Watch my cross section here as I move this around. You can see that it's cutting a section in the direction of that arrow. I want to point the arrow in the direction that I'd like to capture things and just select.

Once I've selected, you'll see that I can move my mouse, and it's not going to move that anymore. We do not want to grab this right here and physically move it. I'm just making you aware of that so that I can hit Done now. And I can delete this because that's what you're going to want to do if you have that happen.

You grab that and move it, it's going to adjust everything that you've set inside the cross section view, a lot of different things. You're almost better off starting over, like, what we're doing here. And that's why I'm showing you in the beginning. But you can take the time to actually move and restructure everything if you'd like. But to me, the easiest thing to do is just hit Done and start over.

So I'll go ahead and start my linear feature extraction again. I'll point in the direction I want to go. Now, I can see my cross-section view here very, very well. First thing we'll do is talk about some of the things in settings.

You do have some, what's called, curb settings. Right now, the linear feature extractions are really only made to extract things like curb. So it's not really made to pull that white line right here for lane striping or anything like that. There are ways to get that, though, using the automatic extraction. I can show you that here in a little bit maybe. But it's really made to generate this curb, which is fantastic.

We're going to leave this alignment algorithm threshold at 75%. If we had a place where the cloud got a little bit-- I don't want to say-- we'll just say less dense, OK, then I can move this percentage down, and it will still continue to try and do automatic extraction. There's other ways to extract at that point too. So we'll talk about that when it happens.

This step interval, it's going to go-- this is just telling it the interval that it's going to extract is going to be every one meter. And then your cross-section depth. You really have to see this in order to understand it. Notice and watch this cross-section view.

As I move this up, you can see that the view gets a little bit thicker. So right now, that section depth is two meters. So it's looking at a whole lot more of this point cloud than what if it was at 0.3. Now, we're going to leave it at 0.3 because we've got plenty of definition here.

If you don't have enough definition, meaning that this looks very sparse in here in this cross-section view, that is one way that you can actually help with the extraction. And that's to bump this cross-section depth up, and it'll fill out this cross-section view a little bit more. So that's what I wanted to talk about. We're going to leave everything alone for this particular instance. I just wanted to talk through it.

Now, next, what we can do is we move each one of these points to the location that we want it to extract a feature line. Now, first thing I always get asked, what if I have more than four points? Well, we're going to be able to add more points.

We're also going to be able to tag these points in order to give them a definition so that when they come into Civil 3D, the features are actually tagged with the name of what they are, which is very, very handy.

So first things first, let's just drag and move these. So I'm going to drag this red point over. Now, I want to show you. This is very, very visual. So I can see that's really close to my edge of my back of curb. Now, I'm going to zoom in here, and you can almost see a change in the color of the points.

I'll drag that over a little more. That's going to be more on the edge of where I want to be. I can grab this second point and move it to the face curb. You can see that orange points now sitting on top of that curb. I can move this point to my gutter. And I can move this point out to my gutter flag if I want, which would be right out here.

Now, I would zoom out a little bit. Man, you can really see where the gutter flag is, in this particular instance. It's right in here. So I can grab that point and move it to right there. That's going to be the point I want to collect is the gutter flag.

Now, say I wanted to add a point for the stripe? The only way it's going to pull that stripe, it's not going to intelligently pull it, it's not going to move in and out, but it's going to pull it because it's concentric down that line. We could add a point.

So I'm going to select Edit Feature Points. You see these grayed out points, you just select one and that gives you a new point. And then you could grab it and move it wherever you want. So I can grab it and move it out to, say, right here. Or I want to delete that point. All right, let's pick that and delete it. I want to move this. I'll move it back. Notice it's on that white stripe now. So we could hit Done here, and we could tell this to extract.

Now, this particular instance, I don't necessarily need that point for the paint. So I'm just going to go in. And I'll select that point, and I can delete it. And then I'm back to my normal four points that I have set.

Now again, I could create as many points as I want across the section view. I can give them specific tags, so on and so forth. How do you give it specific tags? OK, first things first, I've been navigating inside of the cross-section view a little bit.

If you want to zoom, I highly suggest you just click forward on your wheel one click at a time because it's very sensitive. So if I click one click, then I'm OK. If I zoom back out, it's the same thing. Rolling back on my mouse, it zooms back out. If I right click, I can pan. And my left click is my select.

So remember, wheel to scroll in and out, zoom in and out, right click actually pans. So I'm going to zoom back in. It took me a little while to remember that, for some reason. So that's why I like to bring it up.

All right, now, I want to add tags to these feature lines or this line work. So if I select one of these points, I can select this tag. And notice I've already got some tags in here. And it comes up, this is a kind of a long-winded text, this top back of curb. We'll go ahead and select that since it's already here. So I just pick it. It's going to make that top back of curb.

Now, this face of curb, if I have a tag, this is what it was called before. This BOB was something else that was inside of this. So I can close that, and I can do a search for, say, top face of curb. So T-F-O-C. And if it's not there, then I can hit Enter to create a new tag. So I can backspace this out, hit Enter. Actually, top F-O-C, hit it again, and it comes up, and actually, you're able to select it now, OK. So then I can select that tag for that particular point.

This point here is my gutter. I've already got gutter defined in here. So I've got it called out as actually gutter flag. If I wanted to just to say gut, then I could pick a point here, close that actual tag that I was using, and just type in G-U-T and hit Enter.

And it's looking for this gutter flag. So it's going to be a little bit cantankerous here. So if I say G-U-T and hit Enter-- we're just going to select gutter flag, because it's trying to add that. So that's OK. Gutter flag is just fine for what we want to do. Whenever you're wanting to add another tag, you just hit Enter in that dialog box, and it should come up and allow you to create a new tag.

All right, so we're going to add this edge-of-payment tag right here. We're going to call this edge of payment. Notice, that when I put in the search, I type in E-O-P, maybe I've already got that. Now, it says no results, press Enter to create a new tag.

I want to point this out because it happened to me just a little bit earlier in this recording. I was hitting Enter, but I was hitting Enter on the number side, the number the numbers on my keyboard. There's an enter button there.

I have to physically hit the Enter in the middle of the keyboard next to all of the keys that I'm typing with. If I hit Enter there, then it takes it. I have a bad habit of doing that. I've got two Enter buttons on my keyboard. And I hit the wrong one.

You have to hit the Enter button that's next to all of the letters. What you're using whenever you're typing to hit. Hit Enter and return, not the one on your numlock, or your number side of your keyboard if you have one. So that's how we actually add that particular point and that particular tag.

So once I'm done, I can just simply hit Done, and then I can select Start Auto Extraction. Now, I'll go ahead and select that, and it's going to run down this curb. Now, I can pan around and look at this at a little bit different angle, just by right clicking and moving around.

Notice this 90%, 95%, or actually 90%, 89%, so on and so forth. But it says the baseline is 75%. Remember the 75% that we looked at inside of our curb settings. That was our threshold. Well, if it doesn't see anything more than that, then it's going to stop.

Now, I'm going to let this try and go through this. Boom, it stops. And it says, auto extraction is paused. It ran out of information. Notice the number, 53%. That is way less than this 75% that we told it not to go below. That was the threshold.

So we can do one of two things. We can tell it to ignore, and I'll show you what it does when you do that. If I hit Ignore, it goes in a different direction. So we want to stop this. So we just hit Stop.

And if I hover over right here, I can right click, and tell it, I want to delete that selection, which is that set of points that it picked right there whenever we tried to go through without the information we needed. I can right click on it again, and I can tell it I want it to delete that piece. So that gets us back to the last point where we know that it extracted something good. It's at 86% there.

If we want to stop the extraction here, we simply hit Done. So I'll hit Done, and we can see our linear features here start to pile up under linear abstractions. And I can see that it pulled my line work. It did that automatically.

Now, I'm just going to jump across real quick. And we're going to pull the rest of this to this barrier. So very quickly, pull forward. We will just move this to here, here, here, and there. Actually, yeah, that'll work. We'll pull it to right there.

Now, we need to check our tags and make sure. All right, so I'll hit Edit. And I will select my tag. That's still top back of curb. That's not the one we wanted. We wanted top face curb. So we could say, T-F-O-C. We can pick that. We can select the green point. That's the gutter flag. That's what we had before. And we will select our, we'll just call it flag, or it could be EOP.

If flag is not what we want to use, and we do want to use EOP, we can just type in E-O-P and select it. Again, if you wanted to add a different tag or a new tag, hit the Enter button that's not around your number keypad on your keyboard. Once that's ready, we can hit Done, and we can start to extract again.

So we can let it go around this. We can tell it to stop or ignore. We'll just tell it to ignore in this particular instance. It's at 74%. We could say, stop and then have it autorefine, and then start the extraction again.

So we're going to come all the way around, until it gets to this barrier wall. And this is the great part of the presentation in my opinion. I love to show this. We're going to stop this extraction. And we're going to hit Done.

Now we've got a couple of different sets of linear features that are pulled inside of this. Now we're going to pull this wall, this barrier wall. This barrier wall really acts as a giant curb, OK. So we can pull all of that information too.

If we go down to Linear Feature Extraction, and we just select a point, a point in the direction we want to go, notice that we have this wall. Now, so I'm going to pull or zoom back and grab this point here, move it to here. I'll move this point to here, move this point to here, and this point, let's just say to here.

I could move it out. If I wanted to pull it out and put it on that stripe, I could because this stripe does look equidistant from the bottom of the wall. So it's going to pull that information too. So maybe we could do that. If I want to change the tag here, then I would simply go in and edit my feature points, select that point, pick the tag.

This is going to be-- let's call it top back of wall. So let's call it T-V-O-W. That's not there. I hit Enter. Now it's a new tag. I'm going to do the face of wall. I'll select here. And I'll select this tag. And I'm going to call this F-O-W, Face of Wall. That's not a tag we have in this project either. I'll hit Enter. Now it is.

Same thing here, I'll select this point at the bottom of the wall. And we will call this B-O-W. I already have that tag in here, So. I can just select it, call it good. This point here, this is just a point that's inside between the edge of pavement and the bottom of the wall somewhere, I could actually grab that point, create a tag, and just call it-- we'll just call it P-A-V-T, for pavement. Hit Enter, it's a new point.

Now, let's say we did want to add this particular stripe. So we could actually go in here and select Click to Add a New Point. And I'll move that point out. Notice, I've got a pan to the right just a little. And you can see where that stripe is. It's a little bit lighter white. So I can move that out and put it on that stripe.

If I zoom in here, you can see that it does become a little bit wider, and this goes a little bit darker gray. So that's the point that I'm going to leave it at. I'll click on that point and give it an actual name. And we've got one in here already called white stripe. That's fine. If we wanted to create a new one, we could. That's what we're going to leave it at.

So I'm going to back out. And it is going-- I'll zoom back here. Once I've got my points set, I've got my tags set, I can simply hit Done, and then start auto extraction. Now, it's pulling the information down that wall. And it's going to pull it at every foot.

Now, this is what you need to understand. I'm going to stop the extraction. And notice that this does start to veer away a little bit. That means that the stripe may move in and out. Again, this is pulling the curb. That's just one way that we can actually extract that stripe.

So I could tell it right here, I could go back in, and I could edit the features, and I could just select that point and get rid of it. Now, do keep in mind that it is taking it completely away out of the project. But if it's not concentric down the line, there's other ways to pull that line. So we can just get rid of it for the time being, hit Done, and start auto extraction again. And it starts to veer off. So we don't want it to do that.

So I'm going to say stop. I'll right click here where it went off the rail there, and I'll delete that. We'll just go ahead and hit Done here. And if I want to start again, then I can simply go to Linear Feature Extraction, and I can start all over again.

Now, I'm not going to start right here. What I do want to show you is this. I want to show you-- let me go back. I want to show you how the wall changes. The wall actually changes down here a little bit. And I think it's whenever it starts to drop off here.

This wall actually has a different canter to it. So we're going to pull from here. I'll pick this point. I'll turn this direction. And we'll do this a little bit different. We'll take the time to set this up.

I'll move this to the back of my wall, that point there. Face of the wall is here. This point is going to be right here. That's just a point on the wall. If you remember, we called that bottom of wall, and this the pavement. If we edit those and look at them, they're still called that. That's the flag. That's fine. We can call it P-A-V-T if we wanted.

I can select this green point here and look at the tag. That's gutter flag. We don't want it to be that. We want it to be B-O-W, Bottom of Wall. Now, we need an extra point to gather that point right there where the wall comes down and turns out.

So I'm going to select here to create a new point. And I'm going to pull this into here. And then we'll give it a tag of wall point, just W-A-L-L-P-T and hit Enter. And that's just a wall point.

This point here is our top of the wall, top face of wall. So we'll call it T-O-W. Well, there is no T-O-W in here. So well, actually, I think that there is a top face of wall Nope, there wasn't. We'll hit Enter. Now we have it. And then we've got our top back of wall. Now I'll just call it B-O-W. And we've got them set. Now, I'll go ahead and hit Done, and I'll select Start Extraction.

Now, notice it's pulling, where the wall comes down and turns out like this then going out. So our surface in Civil 3D will look fantastic. And it will reflect that particular information right there.

Now, I want to show you what it does whenever it gets to this pole. Sometimes it'll go through those. Sometimes it stops like this. We can just hit Ignore. And if we hit Ignore a couple times, then we probably want to go ahead and hit Stop. And then you can just tell it to extract next here, and it will keep going.

Now, the reason that it wanted to stop was because there was a dark space through here, meaning there was probably a shadow or something. So we could continue to hit Extract Next, Extract Next if we want. We can always tell it to auto refine if we need to, and then hit Start Auto Extraction again. And it picks back up because it can see more of the information now. It's out of that saddle area, but we took something typical and pushed it through there.

So we could continue to let this go down, just as long as we wanted to. We're going to stop it because of our time here. We're going to stop the extraction and hit Done. Now, we've done a lot of work in a not very long amount of time, OK.

We could come back, and we could pull stripes if we wanted. I'll just show you this real quick. I'm going to open another point file in a minute show you another way to do this. But we could pull these stripes if we wanted to.

We would simply do this. We would just do Linear Feature Extraction, pick a point right here, and point it in the direction we want it to go. We zoom in here, so we can see. And I would edit my points, just by selecting. And I'm going to throw that one in the trash can. I'll trash that point and that point and hit Done.

So I only have one point, and I'll move it to here. Now, if I want to pull manually a line or a piece of line work, I can hit my Control button on my keyboard and just pick on the screen, and it will draw that line. So I can actually pick Pick, and there's my line work.

Once I've got that line work set, I can hit Done, and it's done. It's pulled that line. And it's set that line on top of that point cloud. So I could actually manually go in, and I could pull all of these stripes by just using the manual method, the control click method instead of auto extraction.

I'll show you something else that we can pull here shortly. But right now, meaning I want to show you the electric lines, like, on a rail or something like that. That's what we're going to pull here in a minute.

So we're done with our extraction. Now what we want to do is we want to export these. So I'm not going to export all of these. I can. But I am going to do this. And I could actually tell it I want to export them all, if I wanted.

I could just select this button right here. And I can say, I want to pull a land XML file. It's going to output a land XML file, or I could use a DXF file. My method that I like the best is the XML. It works very, very smooth. I'll hit Export, and it pushes that out. It's done.

Now, let's jump into Civil 3D, and let's see what it gives us. So I'm going to open Civil 3D. We need to put it in the right coordinate system. So our drawing, we need to make sure that it is in the right coordinate system. And in this particular case, that is GA83-NF, or is it CF? Not the right coordinate system. GA83, that's not right. Hold on one second, 83-.

Not reading that coordinate system for a minute. Maybe I've got the wrong one. Let's check. We'll go back and see how we can check and see what the coordinate system is. So we can open up, ReCap again. And I will pick that point cloud.

And if we want to check what the coordinate system was, we can simply go to our settings. Ah, GA83WF. It pays to understand what coordinate system you're in or remember what coordinate system you're in. So I bet you it takes this.

We'll go here to edit drawing settings, say GA83-WF. We stone US foot, that is the coordinate system that this point cloud is in. Now we simply want to bring in that XML file. Now it exported that out from the Cloud to my Downloads folder on my machine. So I can simply type land XML in.

And I can pick right here, let's pick the-- go down to This PC, and we'll select Downloads. And here it is, road rehab. That's what it just pushed out. I'll hit Open. Here's all of the linear features I can bring in. I'm going to bring them all in. Why not? I don't have to adjust anything. I can just hit OK. And it brings all of that line work in.

Now, it does not stylize things when it imports it, so you would have to go back and actually set a specific style. But if I was looking at what is that line right there, I can right click on that line and look at the feature line properties. And it says that is EOP.

So remember we tagged it. I can hover over those. Again, that's his feature line. That's gutter flag. This line here is a top face of curb. T-F-S-O-C. and then this should be B-O-C, or top back of curb. I actually spelled it out.