Meshroom Photogrammetry Tutorial (1/3)

in this video we're going to look at photogrammetry and creating a 3D model from a series of photographs that we've taken to create a realistic photo scanned model which can be used within our project so the photographs that I've taken are on a micro scale but you can also look at macro scale and further along in this video we will look at importing the photogrammetry the mesh that we've created in mushroom into 3ds Max to use within our projects some of the tips that I've got for taking images I looked at just a rock and you can see from the photographs from my iPhone it took some pretty high quality images that I found at the park and yeah I think that upon a lot of trial and error the tips that I would give for the photos of the subject matter is to avoid objects which are in the Sun so you can see here that with the images that I used I've used a rock which is sitting in the shade I also tried to keep the photographs that I was taking um as well as I could as I was just doing it by hand I didn't have a tripod to keep the same distance roughly the same distance for each of the photographs and you just want to make sure that you photo 360 Degrees around your object and also taking some photos which capture the top of the object and around the edge to create a really cohesive um uh Point Cloud which will create the uh the mesh of the 3D object so all in all I took 81 photos of this Rock so once you go back into mesh room this is the interface that you'll come across when you open up the app in this section here you've got your drag and drop for your image files that you would like to use the next Black Square here is the image viewer once we've dropped our images in down in this timeline section is called the graph editor then you also have the 3D viewport once it's started to generate your mesh and then fifth we've got the this section here is the nodes and attributes and then you've got preferences this toolbar on the right hand side okay so I'm going to start to drop in the images that we want to use of The Rock so I'll just select the images that I've already organized I'll drop them into the drag and drop area and you can see here that this is the image View um it is important to quickly go through your images that you you have taken and remove any which uh blurry or slightly blurry um so I'll just go and go through that now so this app definitely has many different functions I found the simplest way to achieve what I wanted to achieve to get a 3D scan to import within my own projects I just went down to the graph editor along the bottom here selected structure for motion right click and compute where I'm just going to save it at the end so for the purpose of this I'll just press don't save now you can see here these this is actually a timeline of the process so you can see that it's softer should I have I've actually instructed it to compute up to structure from motion so it does take a little bit of time you'll see over here that you'll get these little camera angles that'll start to set up and this is this is the computer piecing together the mesh from the images that you've provided so once it ticks through you can see that it just um completed part of the feature extraction then it'll work through the timeline and once that is done you'll start to see the outcome in the right-handed viewport okay so that took quite a while for it to go through probably 15 or 20 minutes I suppose it depends on how how strong your computer is so it's just slided in as upside down here so I've just I'm just rotating Around The View but you can see that all of these oh all of these images have created the scene of my object okay so I've just been oscillating around in the viewport so the next step we're actually missing the depth map which gives it a little bit more of its realism so we're actually going to go to the next step which is it's exactly the same process with a right click and compute continue and then it'll work through this timeline again up until this step here so there are a few um sort of preferences that you can play around with here you can down scale the size the amount of cameras basically it's um the it'll really depend on the strength of your computer um I've found that these preferences which are the default preferences for me have worked fine um again it's just it's already moved through the preparation for the scene so it's moved on to the depth map and once this is loaded I will come back to the video if you actually toggle through this tab down here the task manager it actually tells you a figure of how much has been done and you can see a little bit more um clearly how long it takes and whether a step has been successful if there's been any issues this is sort of the task this is the tab where you would find all of that information and how this grid isn't level it's not sitting horizontal um you actually can just turn that off here so great that won't give us motion sickness anymore okay perfect so that's a success it's done we'll go back to the graph editor there is another step so we want to actually create the meshing now from all of these different points to turn it into an actual like an object so meshing same process again compute continue and then it's going to work through these next couple of steps so I've just done this off screen actually but um one thing that I found that once the meshing is actually complete unless you double click on this action it won't show the mesh it was one thing that I wasn't really sure what I was doing wrong when I first did this um at the moment it's on wireframe so you can see the geometry that makes up the mesh but you can go to textured or you can just go to the solid and it'll actually create the solid of the scene and to remove the cameras out of the way you can actually you can take the tracking ball off and you can reduce the point size all the way down in these display settings here and you can we can see all of the the depth map has been applied and it's created a textured mesh Okay so we've done the meshing we've got our solid mesh here we've turned off the wireframe but that you can actually see all of the triangles that make up the geometry that we've taken the photographs on so it's quite detailed I want to see it in the solid excellent now we want to lay our our texture as like a material over the top of this mesh and this is taken from our photos so same process compute continue and we will wait again and see okay it's finished so that was the quickest of them all again I'm going to double tick we're going to we need to turn off the solid so we just hide it I'm not sure what that one does I'm going to turn on oh the Gumball and here we are we've got our texture laid over the top of our mesh pretty good so you can see on the left here in comparison so the real image to the meshed image okay so the final part of this part one tutorial because part two I'm going to look in look into the process or the workflow of importing it into 3ds Max so now I'm going to save as and it saves into the see I've done a couple of other trials here so I'm going to uh I'll just do capitals I'll go okay so that's all saving now and this is where our video ends so I'll see you in the next one

Channel: Dynamic Modelling of Cities

Views: 18,353

Keywords:

Id: KTxEvr6r804

Length: 12min 39sec (759 seconds)

Published: Mon Oct 31 2022