Photogrammetry 2 – 3D scanning with just PHONE/CAMERA simpler, better than ever!

Photogrammetry 2 – 3D scanning with just PHONE/CAMERA simpler, better than ever!


Creating a 3D model just by taking lots of
pictures of a real object? Yes! Photogrammetry is back and it’s easier to
use, gives better results and is still completely free You’ve seen our first video about photogrammetry,
right? A quick recap.
You take pictures of an object from all possible directions, throw them at the photogrammetry
software, which tries to estimate the camera positions an creates a “point cloud”,
a bunch of 3D points that resemble the object To get a printable mesh you have to triangulate
the 3D points In the end, we clean the model a bit, patch
all the holes, slice it and we are ready to print! Well, all this sounds really good and simple,
but in reality, our original tutorial was quite complex Luckily for us, things have changed and a
new player has entered the photogrammetry ring Meet Meshroom Meshroom is free, open-source photogrammetry
software with a beautiful UI It handles the underlying framework AliceVision,
which is the result of cooperation between multiple universities, labs and a French post-production
company The base interaction is about as simple as
it gets, drag and drop pictures into the Meshroom window, hit START and wait for the finished
model However, with Meshroom you can also Augment
reconstruction That means you can add more pictures to already half-finished solution
when you notice in the preview that some area could use more detail And even better, with Meshroom you can do
Live Reconstruction! In this mode, you repeatedly take a set of
pictures, upload them to a folder and they get automatically processed A preview is displayed and you can decide
which part of the model needs more detail You then take more pictures and the whole
process repeats until you’ve captured the model from all angles But before we get to play with Meshroom, let’s go over important steps that you should follow when you’re taking pictures for photogrammetry Your smartphone camera will work just fine,
these two models were scanned with one But if you have a DSLR, that’s even better If you’ll be using DSLR, crank the aperture
to at least 5-6 so that the whole model you’re trying to capture is in focus It’s best to move around the target object
in circles, varying the angle and height after each pass During our testing, we often shot 50, 100
or even more pictures to capture every detail It doesn’t take nearly as long as you’d
expect It’s really important that you don’t move
the object between pictures If there’s people or cars moving near the
edge of some pictures, it’s not ideal, but Meshroom can handle it But try to keep these moving elements to minimum The object should make a significant portion
of each image and you can take close-up shots to capture delicate details If possible try to avoid hard shadows, taking
the pictures outside on a cloudy day is a great way to get even lighting from all sides You can use zoom or even mix pictures from
totally different cameras Meshroom is really great in this regard. Can you make a video instead? Yes, but don’t do it. Although technically it is possible to use
video rendered into individual images as an input for Meshroom, the quality is much lower
compared to a standard image the metadata will be missing and you’ll be inputting
hundreds of images Standard photos are simply better What if you took the pictures in front of
a perfectly white background and rotated the model between the pictures? This sort of works, but again we do not suggest it The results are simply much
worse compared to just walking around the model Ideal targets for photogrammetry have textured
or rough surface Capturing glossy surfaces is tricky, but If
it’s an option, you can cover them with some powder like flour or with painters tape to avoid reflections Ok, we’re now ready to begin the reconstruction You’re most likely going to use the standard reconstruction when scanning objects outside
or when you’re simply away from your PC In this case, let’s assume you already took
all of the pictures got home and now you want to reconstruct the 3D model The workflow is really simple Copy all of the pictures to a folder on your hard drive and then drag and drop the folder or the selected images into Meshroom window Save to project to a destination of your liking, otherwise the reconstruction gets stored in a temporary folder. Now you can either hit Start, but what’s
better is to right click the Structure from motion node and hit compute Actually let’s quickly talk about these
nodes Each one of them corresponds to an important
step in the reconstruction, the individual steps are very nicely described on the AliceVision web page You can manually begin a computation of a
node and it will automatically compute all nodes before it If you double click a computed node, it will
load the result Now that’s why it’s better to start by
computing the Structure from motion node Compared to the full reconstruction, this
usually doesn’t take long at all and you’ll get the reconstruction preview All pictures that got successfully matched will have a green check mark next to them And you’ll see the estimated camera positions
in the 3D view Depending on how well this looks, you can
either compute the full reconstruction or if too many pictures got discarded you can
augment the reconstruction or fine tune the settings which are very well described on
the AliceVision wikipedia The full reconstruction may take a while, the nodes at the bottom will turn green one by one as they complete. You may hit Stop at any time and resume the
reconstruction later If you took a lot of pictures it’s not a
bad idea to let the reconstruction run overnight Once the full reconstruction finishes you
can double click the Texturing node to preview the final mesh You can also right click on any of the completed
nodes and select Open folder Even though the output file format is a commonly
used Wavefront .obj which can be imported directly to Slic3r You’ll most likely want
to do at least some very basic clean-up of the model before printing it The finished model printed from Prusament
PLA Silver and Galaxy Black turned out really well, right? Let’s say you’re scanning something at home You took about 60 pictures and the reconstruction
is going well Except for one area, which you didn’t capture
well, some pictures got discarded and that part of the model is now missing detail With Meshroom, you can simply take more pictures
and add them to existing reconstruction It’s essential that you have not moved the
object between the individual series of pictures If you have moved it, augmenting the reconstruction
will probably be very difficult When creating pictures to fill in a poorly
captured area we suggest taking about 5 to 10 photos And you can try to fill multiple areas at
once New pictures get matched against the full set of photos That means that adding new pictures may even
cause previously discarded images to be successfully matched with the new series Whenever you add a series of pictures to an
existing reconstruction a new branch will appear in the Graph editor Again, you only need to compute everything
up to the StructureFromMotion node which is usually pretty fast As soon as it turns green, double-click it
update the preview When you think you have enough pictures for
the final reconstruction Right-click the bottom rightmost node and hit Compute. This is the original model and this is the
one printed from Prusament PLA Lipstick Red Like two peas in a pod Live Reconstruction is the most fun way to do photogrammetry Click on View – Live reconstruction to open
the setup panel Select a folder to which you’ll be uploading
new pictures and the minimum number of images that will be processed at each step. Now hit Start in the Live Reconstruction panel The first series of pictures should include
at least 10-20 images and should focus on the general shape of the object, let’s not
start by taking close-ups Once you upload the pictures to the selected
folder they get automatically processed and added to the reconstruction There is a small catch at the end You’ll have
re-link the last StructureFromMotion node to the PrepareDenseScene node. Right click the link and choose Remove And then drag a new link from the bottom most StructureFromMotion to the PrepareDenseScene input Now you can Compute the final mesh. Almost all meshes created by 3D scanning or photogrammetry Will have a hole at the bottom Luckily for us, we need a flat base that could
be easily placed on the print bed anyway, so a simple plane cut in Meshmixer, Blender,
3D builder or any other program that you prefer will do Secondly, the scale of the model will be pretty
much random, so don’t be surprised when the model is really tiny after import and
simply scale it up There is one technique in particular that
makes a perfect combo with photogrammetry and that’s sculpting We will make a separate video about in the
future but in the meantime feel free to check the tutorials we linked in the description And here’s the final result printed with Prusament
PLA Azure Blue We strongly recommend to check the Meshroom
wikipedia there’s a lot of information about how to solve some errors you might encounter and what parameters are worth playing with And if you want to contribute, that’s even
better, submit a pull request or contact the developers directly by email So are you convinced it’s time to give photogrammetry
another chance? Consider subscribing if you enjoyed the video
and happy printing

100 Comments

  • Jonathan Lin

    April 11, 2019

    I have an AMD Vega 56. What is a good program that I can use? No cuda for me…

    Reply
  • jesus bastidas

    April 14, 2019

    Props to give james maxwell the respect he deserves

    Reply
  • MagicalPrints

    April 16, 2019

    Holy cow that is awesome!

    Reply
  • Clp4tp Frg

    April 16, 2019

    y si atas un cordon a una plataforma rotativa, de manera que todas las fotos se efectúen exactamente a la misma distancia del objeto?

    Reply
  • Garage Of Diorama

    April 24, 2019

    З дэ принтер, для таких занятий, у тебя, скажем таки ,стрёмный! Мог бы и разориться на нормальный топовый! А иначе нах так с фотками извращаться?! Смени 3 дэ принтер

    Reply
  • dyvanna1

    May 1, 2019

    Statues … how do you take photos of above?

    Reply
  • the raisin of truth

    May 4, 2019

    I love what you are doing. I'm trying to generate interest in this technology with my channel too ^_^

    Reply
  • Ding Dong Drift

    May 6, 2019

    my scan keeps stopping at depth map and won't continue the reconstruction, what am i doing wrong?

    Reply
  • Andrew Carfax-Foster

    May 8, 2019

    how does it work for scale of the objects?

    Reply
  • Benjamin Austnes Narum

    May 12, 2019

    Does anyone know of a way to use several screenshots in order to generate a 3D model? There is no metadata for it, so reference points must come from itself.

    Reply
  • Lior Bar-El

    May 14, 2019

    here is my issue. I took many pictures with my phone and very well lite. for some reason not even one of these photos have a check on it. Is there something im doing wrong?

    Reply
  • Manuel Hauer

    May 14, 2019

    So no Mac support right?

    Reply
  • Transforma Plastilina MX

    May 22, 2019

    Hello and thanks for sharing your knowledge about photogrametry i think is really interesting, I'll like to use Meshroom but can you please tell me what are the minimum requierements to use it, please? I'm about to buy a computer that have Win 10, 8gb ram ddr4, intel core i7 4th gen, 1gb Nvidia grafic card… Do you think I can use it whit a PC with this features. Thank you

    Reply
  • Claas Kuhnen

    May 29, 2019

    I tested meshroom too but I feel it still does not provide the detail level Autodesk recap photo does.

    Reply
  • c-law-v

    May 30, 2019

    Hi Josef, would love to see that 3d sculpting episode! (especially with Blender and 3d printing in mind).

    Reply
  • Sugar Touch

    May 30, 2019

    Any chance for non-NVidia/Metal mac in future ?

    Reply
  • Samuel Snom

    June 11, 2019

    Could I make a 3d model of my entire school?

    Reply
  • stevo728822

    June 12, 2019

    Big problem with this is that you do not have the legal right to reproduce and resell artwork and products in 3D. It would be deemed a breach of copyright and trademark law.

    Reply
  • Demango

    June 12, 2019

    The GH5 is not a DSLR by the way, tiniest of nitpicks, I know.
    Really quite amazing just how much you can do for free these days!

    Reply
  • dreggory82

    June 14, 2019

    This makes me excited to try photogrammetry again.

    Reply
  • ReflectedArts

    June 20, 2019

    Sadly I don't have a NVidia card…fuck me that was the exact thing I needed to scan something…

    Reply
  • Matouš x

    June 20, 2019

    Well i was trying photogrammetry and i was following this video step by step but lots of my photos didn't match but i thought taht it will show me the result even when the photos are not so great but it didn't and i don't know why. Can you write here some tips please?

    Reply
  • alan perry

    June 22, 2019

    what i would like to know is. why is this called photogrammetry 2. when its actually meshroom.???? can you explain?

    Reply
  • Юрий K

    June 26, 2019

    Thanks! Very useful and clear! But why 2018 version? Meshroom 2019 have problems? Here is 2019.1.0 (2019/02/27) https://github.com/alicevision/meshroom/releases

    Reply
  • Athaariq Ardiansyah

    July 4, 2019

    Now… How to let 3D printer prints multi-colored models? Because…
    Photogrammetry + Multi-color output = 3D Photo Booth bussiness
    Right? 🙂

    Reply
  • PrimiusLovin

    July 9, 2019

    The traditional way of 3D modeling real-world objects is time consuming, so an automated, easy and affordable way of using photogrammetry that reduces production times where 3D modeled objects are used is a good thing.

    Reply
  • Stepanenko Viktor

    July 11, 2019

    Thanks

    Reply
  • Jamie Price

    July 11, 2019

    quick question would this work with in game screen shots so you could export a 3d model from a game?

    Reply
  • Wit Owczarek

    July 12, 2019

    😭😭 only for Windows

    Reply
  • raghu ram

    July 14, 2019

    How much time does meshlab took to create the red robot model

    Reply
  • hund 44

    July 17, 2019

    Does meshroom work with amd gpus because it stopped after the point cloud stage

    Reply
  • Kieran

    July 19, 2019

    Would this work with a water fountain?
    Or will the water movement spoil the end image?
    Anyway great technology!!

    Reply
  • Álvaro Nass

    July 19, 2019

    How well explained! Subscribed..I would like to see a next video with a drone capture of an urban space.

    Reply
  • mbcio

    July 23, 2019

    CUDA only… 🙁

    Reply
  • ArloC

    July 23, 2019

    great video for more advanced scanning!

    Reply
  • Bruce

    July 24, 2019

    xiaomi phone lol

    Reply
  • Denis-Carl Robidoux

    July 27, 2019

    That's the nightmare of the Pix4D company who are the greediest company there is.

    Reply
  • Cihan Boztarla

    July 29, 2019

    Noob question!!! How can i put a design on a constructed image? I would like to get the 3d design of a motorcycle and play around with colors etc to see how it looks like

    Reply
  • Sebas Sanchez C-137

    July 30, 2019

    AMAZING

    Reply
  • Tazzi Tiger

    July 30, 2019

    so how do I figure out if my computer can run Meshroom?

    Reply
  • Stephen Foster

    July 30, 2019

    He touches on it in the video, but the single most important thing you can do to achieve high-quality results is to use a good camera. Yes, you can get good results with a Gopro, cellphone or point and shoot cameras, but if you compare a model generated from a GoPro to a model from a DSLR/mirrorless camera using a good prime lens, the results from DSLR are absolutely superior.

    When comparing models between lower-end cameras and high-end cameras, look at sharp corners/edges, photogrammetric reconstruction with lower-end cameras struggle to produce sharp corners, they come out more round. If possible, use a full-frame camera.

    Make sure you move between shots, no standing and pivoting. Take steps in between photos. Take the photos in a systematic way: go low, medium, and high all the way around. For tall structures, you can also scan vertically (take photos as you move up the structure, scoot over a bit then come back down until the whole thing is covered).

    You want to disable autofocus as it causes focus changes between photos and will cause pixel blur when the algorithm looks at different photos of the same feature. Set focus to infinity and leave it there.

    You want to achieve high overlap between successive pictures, both vertically and horizontally (like 70% to 80% overlap). More overlap means more matched features means a more dense point cloud which means a tighter mesh. If you want to know how good a job the software is doing with your photos, turn the textures off and look at the raw surface. The texture hides lots of artifacts.

    Use a fast shutter speed as well to take motion blur out of shots if your shooting by hand and not with a tripod (this is especially important if shooting from a drone).

    Use a polarizing filter to help reduce glare if shooting outdoors. Like hard shadows, the software struggles with how to handle it.

    Post-process your photos to increase contrast in shadowed areas.

    Reply
  • owlbis

    August 2, 2019

    hello, I have a question: should the photos shoot horizontally or can they also shoot vertically?

    Ps: excuse my bad English, it's not my native language, I'm using google translator……. Grazie (Thank's)

    Reply
  • Creatifer

    August 2, 2019

    Stuck at FeatureExtraction

    Reply
  • Tobiasz Stanford

    August 3, 2019

    This is unreal technology. Going to be using this for scanning objects for my college game project. Can’t thank you enough for showing this.

    Reply
  • Hugo Silva

    August 4, 2019

    Why am I watching this… I don't have a 3d printer

    Reply
  • Berham Horadrim

    August 4, 2019

    Which smart phone camera do you recomend ?

    Reply
  • Berham Horadrim

    August 4, 2019

    When using smartphone, do you need to add location/metadata into the photo ?

    Reply
  • Berham Horadrim

    August 5, 2019

    Will this card work ? it has 96 cores CUDA. GeForce GT 240

    Reply
  • hswami

    August 5, 2019

    I have done as u said but surface in 3d object is rough, How do i make surface smoot..? Help please

    Reply
  • B222006 Gamer

    August 5, 2019

    Can you take burst pictures.

    Reply
  • Sam

    August 7, 2019

    Best photogrammetry tutorial

    Reply
  • Dr DR

    August 8, 2019

    Can we use this obj file for 3D animation ?

    Reply
  • Berham Horadrim

    August 11, 2019

    Does the EinScan include the colors of the object scanned? Additionally, inside the video, you are showing the size of the STL and OBJ file from the GOlem and Angel, but is that in MB ? or in GB ?

    Reply
  • hermenegildo sacasonapan

    August 18, 2019

    fucking amazing

    Reply
  • TAMIL STUDIO

    August 23, 2019

    Software name

    Reply
  • basspig

    August 24, 2019

    Tried it last winter. It needs a lot of work to make it usable. Not ready for prime time.

    Reply
  • A New Day

    August 24, 2019

    Is it possible to scan a human being?

    Reply
  • RearAdmiralTootToot

    August 24, 2019

    Now I just need that chick at work to stand still for a bit… maybe if I tell her it's for science?

    Reply
  • Ladislav Panko - Cynicc Art

    August 27, 2019

    Super paradni, obcas se to hodi, ve vetsine pripadu se mi vyplati to vymodelovat v Blenderu 🙂

    Reply
  • Stavros Kyriakidis

    August 28, 2019

    Thx

    Reply
  • Jo jo

    August 29, 2019

    3:45 but that sounds ideal for something like this. Why doesn't it work as well?

    Reply
  • Spencer Yancey

    September 1, 2019

    I have to try this. Thank you so much for making me aware to this process.

    Reply
  • Spencer Yancey

    September 1, 2019

    Lets 3d print more bronze copies of the Confederate States of America monuments!

    Reply
  • Mr Spectacals

    September 3, 2019

    eventually photogrammetry will get so good we'll be copying dollar bills.

    Reply
  • The Annoyed Mr Floyd

    September 6, 2019

    Cosplayers could be photographed for figures this way. Oooh!

    Reply
  • Töffer Orellana

    September 6, 2019

    pls help

    Reply
  • Grzegorz Wronkowski - Vlogger

    September 6, 2019

    "But when you have DSLR…" and showing Lumix GH5. It is not DSLR !

    Reply
  • Творческая мастерская Skott Calderа

    September 8, 2019

    https://www.youtube.com/channel/UCFV3gJcQ2S70p2iuQu5EPEA

    Reply
  • Abeille Magique Dans Coquelicot

    September 11, 2019

    Dannng thoses shoot, so pleasant to watch

    Reply
  • echelon2k8

    September 12, 2019

    8:08 Did you try and fail to make these the same size or did you deliberately make your copy a little bit larger than the original?

    Reply
  • Abhinav Sivakumar

    September 15, 2019

    Isit possible to put an object on a turntable and ensure the base is visibly turning with the object? If not what is the best way to keep the camera still and still scan well?

    Reply
  • iOn-Christopher DiMeglio

    September 16, 2019

    Hmm – well I'd still like to know where that statue is and what it's attached to – does it have a name? (the Deco lioness)

    Reply
  • Foreman cr

    September 19, 2019

    what file format can this software use??? I have tried to use.jpg with no success.

    Reply
  • First Name Last Name

    September 27, 2019

    owo bush

    Reply
  • KarateKeks_Tv

    September 30, 2019

    Hello Josef Prusa I found a solution to use Amd Gpus In the Programm Agisoft Metashape 1.5.5 : First you have to select the Gpu (Tools → Settings → Gpu → select your Gpu.) Now you're done!

    [For the reconstruction you follow the workflow.]
    please pin that on
    This Programm is a little bit difficult But it works xD

    Reply
  • dezo kun

    September 30, 2019

    My camera features a 3D mode. Is there no advantage to 2D footage like u are discribing?

    Reply
  • Cole Burgan

    October 1, 2019

    Are 200 4k images too much?

    Reply
  • ZerqTM

    October 1, 2019

    worth keeping in mind that statues are copy-writable so if your doing anything commercial maybe opt for the really old ass statue who's artist is long dead :p

    Reply
  • S G

    October 9, 2019

    Hi, can I use 3D panorama from my phone with mushroom instead of taking photos?

    Reply
  • Ranit Bhowmick

    October 10, 2019

    Yaa… I once used this… It's kept running for 3days straight and still didn't even finish 50%

    (I was using high quality pictures taken with a really high end DSLR)

    Reply
  • FIORELLO INVENTIONS

    October 11, 2019

    MESHROOM CRASHES… are their any others i could use.

    Reply
  • Geraint Edwards

    October 14, 2019

    Sadly Meshroom does not work on Apple computers! 🙁

    Reply
  • deathknightrevan

    October 16, 2019

    I plan on doing this to my miniatures then uploading them to tabletop simulator for long distances dnd nights. well. trying anyway.

    Reply
  • Galactic Gamers

    October 17, 2019

    Been playing with it for a while now, had many failures. Just realized I need a nvidia card. I just RIPed. Just trying to 3d scan with a 12 year olds budget. lol

    Reply
  • Luis Mendonça

    October 18, 2019

    It does work without cuda, as pointed on meshroom wiki, but you will get lower quality. github.com/alicevision/meshroom/wiki/Draft-Meshing

    But at the time of video recording, I think this was not an option.

    Reply
  • Carlos Rosales

    October 22, 2019

    This was very good explanation of the mesh program I definitely give it a try since it has been updated.

    Reply
  • nicola cornolti

    October 22, 2019

    2:20 "if you have a DSLR" shows mirrorless camera

    Reply
  • Pedro Lessa

    October 23, 2019

    Do you know if works with buildings?
    Or facades for exemple?

    Reply
  • WgoodT - with good taste

    October 24, 2019

    photo…. photogra… photogram…. pho-to-gra-mme-try YAY I DID IT!

    Reply
  • xiang gao

    October 26, 2019

    can i scan my head?

    Reply
  • Rob Estey

    November 16, 2019

    Very cool. Thank you

    Reply
  • thecolonel63

    November 17, 2019

    Such a shame this isn't compatible with AMD. I would 100% be using this for a vrchat room model if it was!

    Reply
  • jameswalker199

    November 19, 2019

    The live reconstruction might be fun if you have some sort of file syncing tool. Take pictures on your phone and have the model already processing by the time you get home.

    Reply
  • Daniel Grove Photo

    November 22, 2019

    Can't wait to play with this! Can I create 3d assets for use in Blender by exporting the resulting mesh and a UV maps texture image?

    Reply
  • Petro770

    November 30, 2019

    You really need to process that mic with some high-pass filter. I hear your neighbor's washing machine doing its dry/spinning program in the background.

    Reply
  • Tangent Arc

    December 2, 2019

    Just when my modelling abilities were starting to improve, this comes along and makes modelling obsolete. I've wasted my life. It is rather cool though. Can't wait to try it!

    Reply
  • fieL

    December 4, 2019

    Thank you for the video.. what 3d printer are you using if may i ask?

    Reply
  • PATOPADRE

    December 4, 2019

    Wonder if Minecraft screenshots work xD

    Reply
  • BATEMAN, O pé rapado

    December 6, 2019

    Does it work to create a 3d scenario like a small room? Or a need to have some "object" in focus?

    Reply
  • Prue Phillip

    December 7, 2019

    This NVIDIA and CUDA stuff is the problem here. The software is amazing but it needs upgrading.

    Reply

Leave a Reply