Messages

1

General / Re: Different camera, same chunk

« on: January 27, 2025, 05:45:00 PM »

it is better to check that it was done correctly (via Tools Menu -> Camera Calibration dialog)

Do you know if there is something specific that I have to check?

See attached, you just need to check that it has created a different calibration groups for each camera/lens combination that was used. It will automatically do this if it detects a different sensor size or focal length, and possibly other criteria. In the attached I have two calibration groups in the top left, though I created these manually as they correspond to two separate but 'identical' cameras. If you do need to manually create groups you can select cameras in the bottom right pane and right click -> create group. You can also merge groups and split groups apart, but all you really need to check is that the groups were created correctly, and if the cameras differ then metashape is pretty good at doing that automatically.

Nothing specific needs to be checked.

2

General / Re: Manually inputting georeferencing data

« on: January 17, 2025, 02:00:43 PM »

In the reference pane you have to double click the cells and then they become editable so you can type values in.

3

General / Re: Export Camera positions and calibration for pytorch3d

« on: December 06, 2024, 05:36:11 PM »

I can't help with the matrix maths, or the internals of Metashape, and I don't follow what you're doing...

But if I did understand I might try using File -> Export -> Convert Images in Metashape to create a set of 'input' images without distortion.

You get a kind of 'pinhole equivalent' version of your input images, where distortion parameters are effectively made zero and the image is centred on the actual principal axis, ready to compare to model renders.

4

General / Re: Create a clean hole in my 3D model

« on: October 23, 2024, 01:56:56 PM »

It was identical to what Alexey said... but ok

I always enjoyed playing spot the difference. This was a tricky one.

• If you are working with the latest release version (2.1.3) of Metashape
• then you can draw the shape in the Model view
• right click on it and set Boundary type to Inner Boundary
• right-click on the model's label in the Workspace pane
• choose "duplicate" option and in Duplicate Model dialog enable "clip to boundary shapes" option

No you draw on the ortho or DEM to make a cut.

You can select outer or inner cut by right clicking.

Hello sonno,

If you are working with the latest release version (2.1.3) of Metashape, then you can draw the shape in the Model view, like olihar mentioned above, right click on it and set Boundary type to Inner Boundary, then right-click on the model's label in the Workspace pane, choose "duplicate" option and in Duplicate Model dialog enable "clip to boundary shapes" option.

Then if you are happy with the result, you can remove the original model from the project and leave only the duplicated copy with the proper hole with the straight edges.

5

General / Re: How to successfully align images from Pix4D Catch app project

« on: September 18, 2024, 08:46:46 PM »

last 2 attachments

6

General / Re: How to successfully align images from Pix4D Catch app project

« on: September 18, 2024, 08:46:19 PM »

Hi jkova,

sorry i was in a bit of a hurry and didn't have time to read or look at everything you posted, but did manage to download and align all your images!

metashape really doesn't like image that are taken moving 'forwards', or images that are extremely similar to subsequent/previous images.

i added all images and checked the 'rotation/yaw' checkbox for all images though i don't think that had much effect in this case, but i did it so i mention it.

i aligned all images using generic and reference preselection, ignoring the warning that the capture distance wasn't specified.

all images aligned and it was garbage as expected.

i then reset alignment for all images. at this point the key points are still present and you can then re-use those to align selected images more deliberately.

i switched the photos pane to small icons and resized it so it was 7 columns wide. 8 or 10 or 5 would have probably also been fine. i then selected a single column of images by clicking and dragging over them, so i effectively select every 7th image.

then i right click and select align selected images. this way it is not trying to align images with virtually identical neighbouring images, and gives it a better chance to work out some proper geometry.

then i selected another column of images, and then repeated and so on, as shown in the attachments below and you can download the project here:

https://we.tl/t-gk5Zxf9oE7

You can align the remaining images if you like, but i suspect that 3 sevenths of all of those images is actually plenty!

7

General / Re: Align Photos Linux Performance 5X Worse Than Windows

« on: September 09, 2024, 01:58:30 PM »

I was thinking of switching from Windows to Linux for my next PC, but this thread got me worried that Metashape might actually end up performing worse even on a more capable machine.

Is this an isolated case or known issue with Metashape on Linux?

Does anyone have any experience of it working just as well on Linux for CPU based tasks?

8

General / Re: Orthomosaik in greyscale

« on: September 06, 2024, 07:40:46 PM »

Looks extremely old.

See page 39 of the Change Log
https://www.agisoft.com/pdf/metashape_changelog.pdf

Dark and Light themes were added in version 1.3.0 in 2016, but the Classic theme has always been available.

I don't know when Agisoft switched the default theme to not be classic, but as soon as they did that I switched it back.

Having used the software since 2010 I already knew where everything was, and re-learning all the toolbar icons seemed like an extra job I didn't need, plus everything with the new themes seemed chunkier and took up too much screen space.

So I expect Dieter is using a recent version, but with the Classic theme applied.

9

General / Re: Underwater Photogrammetry with GoPro Hero 12

« on: September 05, 2024, 06:10:02 PM »

Yes?

Yes

10

General / Re: Photo Alignment produces very strang output result

« on: September 03, 2024, 11:31:48 PM »

Hi DenisIaq,

Here's the completed alignment. Well, 5266/5337 images aligned. The remaining 71 unaligned images are all very featureless. https://we.tl/t-TAv4Oc50p0

The sparse cloud was still very noisy, so I did some heavy model->gradual selection and deletion and used tools->tie points->thin point cloud with a tie point limit of 5,000. I think the noise is predominantly from the white surfaces, and is what causes the 'fuzz' all around it, but the data on the detail areas seems much better.

I don't know if the calibration groups helped particularly, but I am confident that the camera station grouping did help a lot.

Another thought, did you know that you can use equirectangular images directly in metashape? You just have to set the camera type to spherical in tools->camera calibration.

Another thought, you could add the equirectangular images in with the cube face images, and group them in the same station camera groups so they were locked together, and that may help align the 71 unaligned images, because they would only have to align with their equirectangular counterpart to be properly aligned. Then you could discard the equirectangular images if they were not useful for your next stages of processing.

James

11

General / Re: Photo Alignment produces very strang output result

« on: September 03, 2024, 07:39:45 PM »

Hi DenisIaq,

I downloaded your project and started processing with what seems to be a promising workflow, but there are so many images that I'm not sure it's going to finish before tomorrow!

So anyway I can tell you what I did so you can try it if you like.

As Bzuco noted, a lot of the images in Copy of Chunk 2 did align reasonably well, though it's hard to tell because they are dwarfed by a smaller number of very poorly aligned images which throw the scale way off.

I didn't look in any of the other chunks.

I figured that you had converted 360 images to cube faces, so I got an AI to write a python script to set each set of cube faces to a station camera group. see camgroup.py attached. This tells metashape that these image share a 'nodal point' (this might be the wrong term but i know what i mean) and the images are forced to be located at the same spot.

I selected all images and right clicked -> reset camera alignment. Then I tried just aligning the first 50 or so images (select ~50 sequential images in the photos pane and right click -> align selected cameras), and they aligned pretty nicely.

I then selected a load more and repeated, and they aligned nicely to the previous ones.

I did a bit of gradual selection to remove bad points and ran optimise and wondered if it was valid for all images to share a single calibration group as they would have been cropped from different parts of the 360 image and so it might be more appropriate to calibrate them according to which part of the 360 they were cropped from.

The AI struggled a bit with that script, and it's not great, and i didn't spend any time improving it, but it works for now - calibgroup.py attached.

I don't know if that actually makes much difference, but it assumes that all images suffixed with _0 are from the same part of the 360 and same for _1, _2 etc and creates calibration groups for each. you can see the calibration groups in tools -> camera calibration.

I then continued aligning additional batches of images, occasionally doing a gradual selection and optimise, until i got up to ~600 aligned images, with still most of 5000 left.

At that point I got it to align all the remaining images, and that's what i'm waiting for now.

As Bzuco points out it's not an ideal camera and the white surfaces won't help much with reconstruction, but the alignment might come out ok.

I would also suggest massively reducing the number of photos, at least initially just to check that you can get a good alignment. I think you could just keep every 5th set of cube faces so you have more like 1000 images and you'd still have enough overlap for alignment.

I'll let you know tomorrow if the alignment completed this end, and share the result if it's not awful!

12

General / Re: Underwater Photogrammetry with GoPro Hero 12

« on: September 03, 2024, 01:54:02 PM »

I am using GoPro Hero 12 as I mentioned on the caption.

Any information will be great for me.

1- I was capturing a video with wide settings and Hypersmooth "on". I made a research and it came out that when Hypersmooth is "on" It plays with the pixels and it is bad for photogrammetry.

Hi!

Here's a couple of other things you can try if you aren't able to reshoot following some of the the wise and practical suggestions above.

See the screenshot attached.

If you haven't already, take a look at your calibration settings by going Tools -> Camera Calibration.

I believe that GoPro 12 in Wide mode gives a fisheye image so check if you can get better results if you change from the default Frame mode to Fisheye, then come out and optimise the chunk.

Another thing you can try is to right click the calibration group(s) in the left hand panel and split them, then come out and optimise the chunk.

Splitting the calibration group means that each image will get its own calibration parameters, rather than one set of calibration parameters for the whole group.

Hypersmooth must necessarily crop every image, and by different amounts and in different places, amongst other corrections, so a global estimation for focal length and principal point will not be applicable to most, if any, of the images, and this would certainly lead to a bending effect.

Splitting the calibration group is not normally recommended because images with fewer tie points could end up with their parameters being estimated wildly wrongly, and you would normally go to lengths to ensure that your images do have the same internal parameters, but with Hypersmooth it's guaranteed that they won't have, so this may improve things.

There's a good chance that the alignment ends up worse after this but I work with appalling images all the time and can normally wrangle something out of them by trying this sort of stuff.

Also do you have any calibration while using the Agisoft Metashape? Like focal length and pixel size?

It can help to have a precalibration xml file if you have an idea how your camera should be calibrated, and if the images aren't good enough for metashape to figure it out during alignment, but in your case there is no single good calibration because hypersmooth will have messed with each image differently.

Finally, dont go down the route of thinking camera calibration will fix your alignment:

But dare to dream that it might. Admittedly the ideas above are just the tip of an iceberg and there is a rabbit hole down the middle of the iceberg which may not be the most productive place to explore.

13

General / Re: Banana Effect and "Not Enough Reference Data" for GCPs

« on: August 23, 2024, 02:47:07 PM »

i had a similar issue with a single long strip of 200 images yesterday. i was able to add about 8 gcps but these were more or less colinear so not useful for orientation. i was also fortunate that all cameras were facing directly down the z axis with no appreciable roll/pitch/yaw, so setting rotation values to 0,0,0 for each camera (which i think corresponds to looking down z, with x and y coincident with the xy of the image plane), setting rotation accuracy to 0, and ticking the box next to rotation for each image, and optimising seemed to do the trick.

if your images are differently oriented then it might be tricky to work out the required rotation values, but not impossible.

might be worth a try anyway.

Why are you using local coordinates?

i too use local coordinates for almost everything, simply because my work is rarely georeferenced (or if it is then it's to a local projected system) and xy meters are easier to get my head around and record than lat long, especially as most of my work is related to buildings which either have their own local project coordinate systems or the specific coordinate system used isn't relevant to the project which is normally to just get a good image out at the end.

14

Feature Requests / Re: Labeling

« on: July 18, 2024, 05:18:11 PM »

Model -> Show/Hide Items -> Show Labels

You can also assign a shortcut key in Tools -> Preferences -> General/User Interface/Shortcuts -> Customize...

I just set it to Shift+L on mine!

15

General / Re: Information numbers of "selected points/total number of points" has disappeared

« on: July 18, 2024, 01:59:06 PM »

Model -> Show/Hide Items -> Show Info