I suggest using a rotating base with the object sitting on it, your camera mounted on a tripod so every image is at the same distance. It makes the calculation much easier and you will get better results. There are plenty of You Tube videos that show you how to make one.
Also, an idea, you can capture a video of the object, then with any program that converts video to image, and load all images in the software. I will try this sonn enough.
so awesome that this technology exists to do at home! I'm experimenting with 3D scanning and picture digitization, for beta-testing at home videogame development demos.
Great video. I am focusing on a Lower limb prosthetic project for my postgraduate University degree. Due to the current social distancing restrictions I am unable to obtain a 3D scanner so this works as a really good substitute towards my research. Thank you for the insight!
Thank you! I'll have to give this a try to see if it works for less bumpy stuff, like car model kits, to have something I can 3D model widebody kits and the like onto.
3 года назад
Gerçekten daha 3 gün önce 3D obje taraması üzerine konuşuyorduk, scaner yazıcılar gibi bir obje tarayıcılarda çıkar yakın tarihde derken sizin bu videonuza rastladım. Teknoloji çok hızlı.
Here is an idea... If the table were a grid, the program could be made to recognize the grid to give the program a constant reference for distance and angle. Plus it can automatically eliminate the table from the object. ;) Mags
This is VERY cool! :) I have an old motorcycle engine (and some other parts) from around 1910, and I will be selling them soon to pay off one of my credit cards. It would fantastic to be able to create a 3D model of them to save for future reference, because I will never be able to find these parts again. It would ALSO be great to create scans of some of the abandoned railroad structures around here, because they are starting to demolish some of them :( I could have someone print out the buildings in 1:87 scale, and I could add them to my HO-scale railroad layout. It would be a way of "preserving" the buildings into the future :)
@@Veni_Vidi_Vortice I'm thinking a white, featureless background and a mirror as the turntable top. It might reduce the artifacts. And some kind of holder for the camera, steady and consistent.
Love learning about 3D scanning. I hear about this tech a lot, at my 3D scanning company, cool to see how it works. Word of caution: b4 spending $$$ on downstream application like a big print or cnc mill take the data into a separate software package to inspect. Ive had many clients come to me AFTER a disaster costing thousands. Cheers!
I dont mean to be off topic but does any of you know a trick to get back into an instagram account?? I stupidly forgot my login password. I would appreciate any tips you can offer me.
Back in 1995 we could lazer scan an onject and create an IGES file to transfer into a freeform CAM file to machine the part or to create a file to 3D print.
great video. I have tried this 2 years ago using the Kinect 2. I used 3D builder from Windows SDK. This was not very successful. Will try using your software recommended. Again great video. My computer is an Intel I7 with Nvidia GPU. Should be fun
Thanks for the information!! Awesome!! I have a bayonet for a Mauser. The bayonet does not have its grips on it. Long since lost to history. I do believe that I will use this with my note10+ to capture the handgrip of the bayonet and then model in the missing grips and then print them in NylonX. A little epoxy and a couple of brass pig bolts should finish nicely. Not historic, but instead with a modern touch it should look pretty nice.
Mình mới xem qua cảm thấy rất bổ ích, ở Vietnam rất ít chia sẻ kiến thức về công nghệ này. Hi vọng bạn sẻ tiếp tục chia sẻ nhửng kiến thức hay như vậy, cám ơn bạn!
@Long Truong Ke dạ có phần mền đó anh. A có thể nhìn kỹ ở video ak anh. E củng đang tìm hiểu. E mê mấy con 5 trục lắm. Hi vọng cùng đam mê có thể làm quen vs anh ạ
One tip that Prusas video taught me was to right click and compute structure from motion...then look for holes and add photos before wasting time on a full calculation (vs using later portions in the pipeline). This videos awesome, I also use mesh lab to clean up my scans. I did a video on animation as well using mixamo if you scan an action figure which is fun. Great video and run through I’ll have to check out your Kinect versions also!
How does this process handle scale? I can see this process a HUGE help in designing objects for prototype and other projects. As an example, if I needed to design a custom bracket for a fog light on a car. Having scale models of the light and the area they are to be mounted in would greatly improve the ability to create a custom bracket or mount. Scale is important.
Yes, this is actually a very good point. Might be good to put something into the scan area like a cube with a lot of features that has a specific size to scan it together with the model. Then use this to scale the final model to the "real" dimensions. Just a quick thought.
What's the point in adding light sources to get even lighting when you step in front of them when taking pictures (making shadows). :P Any ways, cool video!
Thank you. I have a question. How to set the Coordinates (x, y, z). I mean the default coordinates are not aligned with the height, width, and breadth. If we want to change the x, y, z by ourselves, is it possible?
Great video. Thank you. I have a friend who scans real people with a $35,000 portable hand-held scanner. Super expensive. He prints out the files to 1:48 scale and makes little people for his model railway. Could this camera based Photogrammetry work for scanning people in different poses? Like a workman in overalls leaning on a shovel?
Great stuff.. I am wanting to make car parts for my 80 series landcruiser that are expensive to purchase or cannot be found, can this be used for things like door handle screw covers and clips etc.. also would it help to use a table similar to a lazy susan to rotate the item or do you need to move around the item itself? Not all of us have those expensive light setups so can natural room light work?
I'm going to give it a try. I have wanted to do this for such a long time. But every other workflow seemed to be a bit to complicated. So it didn't get me hooked up.
Neat, I believe it would work. Just have to make new angles, as in have to move the fixed camera to change the angle a little bit after every few shots. Edit: 3:33
very fine quick and understandable video of photogrammetry, thanks a lot. i hav´nt seen all of your videos yet, did you show where you put your 3D Object into a CAD program for editing it?
Maybe a dumb question but after you scanned and got the 3D object, can you use that to 3d print the object? And if you can how would you know if the object is the size you want?
Looks very promising compared to the more expensive rgbd cameras on the market. Is it possible to get real-world measurements of the object from the generated mesh 3d model?
Thanks for the help! Great tut, saved me hours of research. I managed to add my specs to the db file. Which is now located in a different location. Anyway it only took 25minutes for me with the rtx3060!
Is there a totally free 3D scan app? One that you dont need to sign in, no payment after 5 scans, the rear camera works too and works with out lidar ? Im sure there is many people that wants to know this.
If you want too. Personally I found a odd guide on photogrametry online from 2021. Haven't delved too deep into it yet. Otherwise tho I recently just got a kinectv2 with cable with link cable and been looking up the in and outs of what I can do with it since it arrived. Shoulda been doing that before I bought it but oh well.
One semi interesting thing is that there are allot of open source software options for photogrammetry these. Even some good ones that doesn't need a Nvidia donation to function. However it seems they might lack gpu acceleration last I check. Which can obvious be a blessing for hardware compatibility but baloons processing time
Excellent video. I have a question regarding what we are able to do with the image once it has been rendered. I am a teacher and my objective is to create 3D renderings of the sculptures that my students create. I want them to use photogramety to do this. Once we have the rendered images I want to be able to place the on display in a virtual gallery or showroom. I want to be able to create the gallery. Then I want to be able to create pedistals within the gallery as well as have control over the lighting in the space as if one were in a real gallery or museum. Then I want to be able to import the 3D renderings of the students work into the virutal space. Do you have suggestions on how to do this? Thank you!
I'm looking for the best quality scanner not super expensive that will work with an iphone 12 mini or a 2014 macbook pro? any suggestions? looking to scan vintage action figures
How's it going.. Great video.. Was wondering: Would I be able to use this method to create a 3D image of an rc car body to be printed on my Voxelab? the body shell is hollowed and not a solid structure as to why I'm asking.. Tried using Fusion 360 but way too complicated of a process. Thank you..
Hi Daniel: This may be a stupid idea, but can you, instead of taking a bunch of photos, film a video (full HD or 4K) and using a video editor (or Blender) save the image sequence to load into Meshroom? Thanks & greets from Argentina. Marcelo Miguel Bazan.
Why not use a couple of ring lights and a green or contrasting back ground to make removal of any artifacts easier? Put the subject on a turn table to control the angle.
Video capture does indeed work as a decent method of capturing low-resolution 3D geometry. Advantage of video capture is being able to pick and extract a large number of usable frames to feed the photogrammetry software. Disadvantage is that the highest-quality video capture hardware is currently capable of capturing only 4K resolution footage (approximately 12-megapixel stills), when typical photogrammetry software prefers minimum 20-megapixel images for medium-quality 3D scans.
Not really happening. A camera with 0.3mp resolution (640x480) can roughly map its environment for augmented reality but never create a decent and high resolution object of anything. That is basically Xbox 360 quality and you’ve seen how those results look like in the video I did.
Would it be possible to take pictures from an object that is on my Monitor from a game and 3d print it? I have complete control over moving and lighting that object.
Are there any secrets for really small objects? I need to replicate some parts for a 1/35 scale tank model and have tried a number of times with my SLR and each time, the parts are just a blur! They’re no more than 1cm square and look a bit like a three bar gate, so they’re not solid either...?
