Photogrammetry Puns

I need a photogrammetry app for Mac M1 (late 2020) that is simple to use, I tried regard3D but it said that my focal length was wrong, could you help me fix it or recommend another app or program.

PS: I used an iPhone 6s for the photos, would a Cannon IXUS 185 (handheld camera) work better, and worth another try with.

Does anyone have experience with smaller scale 3D scanning or photogrammetry?

I am working on creating a database of digital objects for my Museum. We do most of our designs in 3D (SketchUp and Vectorworks) and have been creating the objects from photos. Throughout the pandemic this has really been helpful to keep our exhibition program moving forward. Our curators are still mostly working remotely and being able to do virtual walkthroughs has really helped solidify things so we're not doing as much rearranging during install.

Generally, our process is:

-We take a photo of the object (sometimes existing, sometimes a new photo from a specific angle), crop the background away and save as png

-In Photoshop, create a path from the edge of the png. Then export the path to Illustrator.

-In Illustrator save the path as a dxf/dwg file.

-Import the dxf/dwg in Sketchup. Scale the shape to the right size, explode it and use it to create a surface that can be extruded. Then we apply the png as a texture to the face.

Now we have an object that looks like it belongs in MineCraft. It's serviceable, but kind of a time consuming process. I'd like to find a handheld device or even phone software that can scan and provide a file that we can import to SketchUp or Vectorworks.

Hello all.

I have a DJI Air 2S and want to try scanning large landscapes. The area would be quite big, maybe a square mile. There's a large lake and surrounding mountains I'd like to capture.

My questions:

1. What flight path is best? Circle? Square? Rectangle? Random lines that match up? And what angles? It seems like you can pretty much do whatever as long as you've covered it all, right?

2. Could I then come a lot closer and take random photos to give better details. Say for example a bench by the lake, or a carpark etc, and then combine these in to the final image so it's better quality?

3. I use a Mac, what software can I use for free that's simple and doesn't require messing around with directories this, and coding that, etc.

4. How does one exactly photograph the inside of a building for 3D scanning? Would I just move forward, take a photo, look up, left, right and down and take photos, move forward again and take a photo etc, then do it all behind me?

Cheers!

Hi, i'd like to get into photogrammetry for some personnal projects. I mainly want to scan large surfaces to find, for exemple, the general flatness and then use it to do some CAD. So accuracy is more important for me. I want to use photogrammetry as a mesurement tool more then anything else. While Googling, I found that photogrammetry could be a cheap solution for me since you need a camera and free software like meshroom.

That said, before investing alot of time to learn everything, is photogrammetry a good solution for me.? I'd like the feedback of experienced folks!

Thank you in advance !

I'm just getting into photogrammetry and experimented a bit with Polycam on my iPhone. The results were better than expected.

I just placed the item on a coaster and started rotating. I did a small mark on the coaster so I knew when I made a complete rotation. N o special lighting. https://poly.cam/capture/32C09EA0-AD78-4375-8382-3E330DD81F0A

I'd like to start taking better photos (probably still with my iPhone). I'd like to improve lighting and possibly rotating the object.

Any recommended light boxes and turntables? I'd need a turntable that either rotated with each click of the camera... or perhaps a button that rotates a specified increment. Should say, I'd like to scan some chess pieces.

Hi.

I design, implement and experiment methods for generating immersive VR environments from convenient and low cost image acquisitions: videos taken by helmet-held 360 camera.

Currently, there are 7 scans for 3 kinds of VR headsets, ie 21 VR applications that can be downloaded at https://maximelhuillier.fr/

See also https://sidequestvr.com/user/330664 for screen shots and video presentations (taken by a standalone Oculus Quest) and installation using SideQuest.

In short:

- The environments are towns (snowy, or semi-medieval ones), forest, Verdon's canyon, basalt canyon, campus.

- Supported VR headsets include standalone Oculus Quest, headsets compatible with Windows Mixed Reality, headsets compatible with OpenVR (in Steam VR).

- I generated the VR applications during years 2020-2021 using Unity 2019.4.

- I used a Garmin Virb 360 or a Gopro Max 360 for the video acquisitions.

- The lengths of the acquisition trajectories are between 670m and 5.2km, mostly by walking (but also biking or kayaking).

- There is no manual clean up (see the "true" result).

- The download sizes are between 143Mo and 777Mo; it can take several minutes.

- Cybersickness can be reduced by following instructions available from the Sidequest page above.

- More details on the used methods are available in my web page: http://maxime.lhuillier.free.fr/

Last I only checked the VR applications on

- an Oculus Quest (two modes: standalone and linked to PC), and a Go

- a Lenovo Explorer (without and with SteamVR).

I hope that one of them will work for your headset.

Hi,

I wrote a script to automate the process of downloading each shoes 360 Image Viewer images and use them as input for apple object capture api. Too bad the soles are not captured in the image viewers, but still the output is amazing. Have a look at. Cheers.

https://preview.redd.it/pdbkll8tz9881.png?width=628&format=png&auto=webp&s=a5266e1c5e8b6befbf34831f3b23fb784517b04d

https://reddit.com/link/rqemug/video/gddybox60a881/player

I'm looking to upgrade my graphics cards to get better performance using CAD software and also for photogrammetry. Specifically using 3DF Zephyr. Right now I'm running two GTX-690s giving me 6,144 CUDA cores at about 1GHz, and of course they are also an older gen of CUDA.

I'm looking at adding an RTX 3060 but now I see they have come out with these LHR (low hash rate) cards to keep the Etheirum and bitcoin miners from buying all the cards.

If something is gimped for mining, is it alsol gimped for photogrammetry?

I would be adding this card as my primary and having my two GTX-690's as secondary cards to upp my CUDA cores available for Zephyr.

Do I need to get a non-LHR card?

As I ask in the title above I was messing around with photogrammetry, and the idea is to scan my head or my arm and have the scale needed to 3D print a helmet or a bracelet, for example if I use my head and I measure my iris then after in blender I can use a sphere with the same measure of my iris and scale the entire model to the scale of that sphere to match with my iris.

The question is could I use to photogrammetry to have a 3D representation of my head or arm and design with those metrics? Or is super inaccurate?

I'm considering a new phone and I figured I might as well look to see if there are any with a depth sensor and if there is any software that uses that sensor to get reasonable 3D models?

If so, can you get good results with it? Or is it too compromised to be worth using?

There are not that many phones out there with a depth sensor so not sure if I should bother.

Hey, i painted another immortal when i was in "kill team mood". i also made a phot scan if a miniature so you can include it yourTTS games. C&C welcome.

Let’s say I’m working with items the size of a shoe, and I’m using a turntable, and the lighting is perfect, and my camera is an SLR with a good lens, and I’m shooting from a tripod, and I can take 1000s of photos. What’s possible? Could I get 1mm resolution? 0.1mm? 1cm?

Before I build an elaborate photogrammetry rig, I’m curious about what kind of results results I can expect for my effort.

Building upon last week's whitepaper on why drone photogrammetry needs a rethink, today we are publishing a technical write-up on how to achieve the vision of a New-Age Photogrammetry.

Technological advances in the last few years, especially the improvements in 3D deep learning, distributed cloud computing, and embedded hardware can alleviate much of the pain points that exist with the photogrammetry software today.

Our technical whitepaper dives deep into how we can incorporate these technological advances in the photogrammetry pipeline to create a smooth, managed, and superfast experience for the end-user.

Building such a New-Age photogrammetry will need the active participation of the entire photogrammetry community, the industry, and academia. We at Preimage are one of the many that are trying to modernize this industry.

We would love to hear your thoughts and suggestions on our thesis. If you want to test our Beta version, do contact us. You can watch the demo of our point cloud classification here.

It is a position that is set to start as early as possible next year in Cape Town, South Africa. The job will be to travel to geological outcrops around the country, build a workflow to collect images (normally via drone), generate models, and then be able to make interpretations of the models and generate virtual field trips to the sites.

We had a candidate selected for the position, but he has backed out at the last minute. As a result, we have funding set to start in January, but nobody to claim it!

PM me for questions and details. Thanks!