How To 3d Scan With Your Phone | Photogrammetry – 3d Scanning With Your Smartphone (any) Camera


Subscribe Here





Photogrammetry - 3d Scanning With Your Smartphone (any) Camera


In this video, I’m talking about 3d scanning with just your smartphone camera and we’re starting right now. Hello, my name is Daniel. Welcome to the Crosslink Channel. I would like to help you being more successful with 3d printing and if you here for the first time, subscribe and hit the bell notification. So you don’t miss anything and the last three videos that you can see on the playlist here? I’ve been doing 3d scanning with the note 10 plus and then with these two sensors, the Kinect for Xbox One and the Kinect for Xbox 360 and today we’re talking about how to do 3d scanning just with a smartphone camera and some additional software on your computer, so let’s get started. So the idea about photogrammetry is that you take a lot of pictures of an object from all different angles around the object in multiple turns and multiple angles and then you pass it to a computer software that is called a photogrammetry software and that’s going to calculate a 3d image from all of those images. Now there is, of course, some prerequisites for this to be successful. You need a camera and you can use a smartphone camera and it can also use other cameras we’re starting with a smartphone because that’s what everybody has in their pockets so to do all the photogrammetry magic, There is multiple open source and free bare solutions out there and we’re going to focus on mesh room today, which is one solution that you can use. I’m focusing on that today because it’s the most simple approach. You can just pass the software your photos, and then you can hit basically one button to calculate the final 3d object. You can customize every single step in the process, but you can also just use it as it is and get some first very early results very quickly, however, one prerequisite to do That Is you need a computer, which has a GPU card more specifically. You need a card that supports CUDA, which is only supported by Nvidia graphic arts on the other hand you it also needs to be a powerful PC. Otherwise, you will have to wait for quite some time for the results to come, so my computer, for example, takes about one hour to calculate a final image from roundabout 5280 pictures compared to the sensors here, which basically gave you immediate results. Also the No 10 plus took just one minute at Max to calculate a finer 3d object, so that was pretty quick, but the results were also mixed. I would say, so let’s start by taking pictures of one of the sample objects that I’ve been using in my last videos. The lion statue. So here we go. This is the lion statue that I want to convert into a 3d model using photogrammetry and the first thing that you need to make sure is that you have good lighting, so I’ve a light here. That’s shining from the back of this object, so we have some light here from this side coming. Then we have another light coming from the other side, so we have lighted up the front evenly and you can see these are a few slides, so it’s not harsh light, hitting the object. That’s one important thing. Now we have a final light source coming down from the ceiling to the object, so we have some decent light from almost every angle, and our object is really well lit, so the next step to do is take pictures from every possible angle of this object, so we need to step back a little bit get to get the object in the frame, make around about 40 pictures from an angle that’s slightly above the object and then go down and take another 40 pictures from a lower angle and then finally do some close-ups, for example of these dark areas here to catch all the details that might be hidden in the larger images because these areas being a little bit too dark and once were done taking these pictures were going to pass these pictures into the Photogrammetry software, so let’s head over to the PC and take our pictures and convert them to an actual 3d object of list line statue. Okay, so I’ve opened up mesh room now. And this software can be downloaded for free from the website. I’ve linked in the description down below. So what you want to do? Is you want to drag these images in to mesh room into this area where you can say drop? So we’re doing this now? These are the images that I’ve just dragged All These images have a little icon here on the upper left corner, which means it detected the right camera. So you have to make sure that the images are done with a camera that this software knows it might be some exceptions where the software might not detect your smartphone and that could be an issue because it needs to know what kind of sensor and focal length the camera is using and that’s important for the software to know for doing the right geometry and calculations, so this works what a Google pixel and probably most iPhones will also get detected. The next thing I want to show you. Is this bar here where you have all these different steps in the process that you can individually select so when we just want the 3d object and not interfere with any of the settings here we will just right, Click the texturing box and then hit the compute button. And it’s gonna ask us if we want to save it, so I don’t want to save it for now. And then it’s going to start calculating the 3d objects, and it’s gonna start here at the upper left corner. You will see a progress bar, which is going to move slowly from left to right until the whole process is done. And that’s taking quite some while so on this computer. It’s gonna take probably one hour. I don’t want to make you wait so long, so I have done this already and I’m gonna load the results for you now. So what I want to do Is I want to show you Some of the intermediate steps involved in the process, so the first step that I want to highlight here is structure for motion, so what it does is calculates a 3d structure in terms of generating pixels in a 3d space, so you can already have a good impression of what will the object actually look like it’s not yet a complete 3d structure. There’s no vertices yet, but you can already guess whether the result will be something decent or not the next step. I want to highlight here is the meshing step and the meshing step. Is this one here? And what if we switch over to the meshing step and double click it, you will see the results from the meshing step, so I have two like this one here, and that’s the result from meshing steps so here you can see the final mesh, the 3d structure and you can also see that some garbage that has been created around this object, so these are all things that are probably in the room around the object that have been captured by the camera, somewhere in the background and software still tries to convert all the things that it sees into 3d objects, so what? I’ve done here to clean up the mess a little bit. I’ve narrowed the angle of the camera so to say a little bit and went with this variant of the step, and that shows me. Okay, that’s already much much cleaner. There’s less objects that are actually not supposed to be here. If you look closer, you can see that the result is really very fine, detailed, and you can alright also see that in the darker areas of the image, we might have some issues that we need to fix like here underneath this. Ya shield of the lion. And also the chin area seems to be a little bit weak and here in the darker areas, as I said, this might be some issues that we need to fix by taking more close-up pictures and adding them to the list of pictures here on the left side and that might improve the result in these specific areas, but overall. I would say the amount of detail already from this first. Try is pretty nice, and then we can switch over to the last step, which is the texturing box and texturing step. And if we double, click this so here in this last step you can see. This is the final mesh, overlaid with the textures from the MHS, and that looks really pretty pretty nice, so I can imagine this could be printable, However, what we now need to do what this software cannot do for you. We need to basically remove everything That’s not supposed to be in the 3d object with another piece of software, so I’ve loaded the object in mesh lab, which is another freeware, which can be used for editing 3d objects. And you can see, there’s a lot of things in this object that we don’t want. And how can we remove it? So it’s pretty simple in this software, You just have to use a selection tool and select everything that you don’t want and then basically hit this delete button and that it starts removing everything that’s not wanted so we can continue with this and just remove everything that we don’t like so, and I think we’re pretty much done with this objects What I’m going to do is I’m going to put these objects onto my website, link it in the description of this video, so you can download the object and have a look at the final results, Okay, so some find the words on the mesh room software and photogrammetry overall? I’d say you should have seen that. This method is really powerful. You can take almost any kind of camera. Take pictures of an object and then run it through the software to get a 3d object. That works pretty well. However, in some cases, you might have to try different lighting situations. You might have to make sure that you have the right distance, so you also need to make sure that the pictures are really sharp in every situations that your lighting situation doesn’t change all the time. These are some of the parameters that might influence the quality of the outcomes on the other hand, the results you probably have to post process the free objects with a software like mesh lab to remove parts that you don’t like to close gaps to close openings that you don’t want. So it’s some experimentation involved, but overall. I’d say with mesh rooms You can get from the pictures to the 3d objects pretty fast as I said, you need a pretty beefy computer and then one of the following videos, I’m going to try out taking pictures from a drone and then converting these into 3d objects are with something like a building or anything bigger, So let’s see how the results are from that and yeah. I hope you’ve got some fun out of this video. If so, like it, subscribe share all the good things, So have fun 3d printing. Bye bye. [MUSIC] [Music]!

3d Printed Master Chief Helmet | 3d Printed Halo Helmet

Transcript: Hey, how's it going, guys? Just, uh, thought I would share with you. A project I've been working on. This is my master chief or your halo mark 6 helmet. And this was 3d printed on my ender threes. Uh, so I've got an Ender, Three and Ender, Three pro. And,...

read more