Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Topics - Kiesel

Pages: [1]
General / incomplete point selection in point clouds
« on: November 23, 2022, 10:49:24 AM »
Dear Agisoft team,

I have difficulties with the current selection tool for a long time, especially when I work with deeper point clouds like imported laserscans. The current selection tool seams to be restricted to depth, so I have to select again and again the same area to a deeper and deeper space up to ten times and more and even then I can't be sure that I have selected all points in this area, because when I zoom out some "new" points arise. And in this outzoomed view it is of course more difficult to make a precise selection.  Have I overseen something? I hope for a solution soon for an otherwise nice product.

Best regards,


General / Problem with using depth maps of Google Pixel smartphones
« on: July 11, 2022, 03:31:30 PM »
Dear Agisoft team,

Today I tried to follow your article with photos made with the Google Pixel 4a 5G, but when I import those photos with 'File/Import/Import Laser Scans...' and then choose all photos with 'Dynamic Depth Image (*.jpg *.jpeg) after that Metashape opens surprisingly the same folder again, which now seems emty, to choose a folder. When I go one folder up to choose the same folder, with the photos in it, I get an error message (see appendix) that Metashape can't read the embedded depth data. As result none photos are imported.

To make sure that the photos have embedded depth data I have imported one in Photopea, with the result, that they have 4 layers (see also appendix)
Layer 1 white filled depth data layer
Layer 2 depth map
Layer 3 original photo
Layer 4 with depth data processed photo

In advance to any help

best regards,


Feature Requests / Use of vector coded targets
« on: January 31, 2022, 05:47:17 PM »
Dear development team,

I would like it if you could add the functionality of vector-coded targets to Agisoft Metashape. Sometimes it is necessary to signalize corners, which isn't always possible with normal or coded targets, but it would be with vector-coded targets as they are used in other photogrammetry packages.
For an illustration of what I mean, I have generated some by hand, see attachment (tip of the vector-coded target is in the half distance of the two targets).

Best regards


General / Import of Leica BLK3D data, automatic model orientation?
« on: July 09, 2020, 11:13:21 AM »
Dear Agisoft team,

I have successfully imported Leica BLK3D data and generated a scaled model out of it, without additional measurements.
So far so good, but my model doesn't have the right orientation and large errors in m and °(see attachments).

Is is also possible to orient the model by the measurements of the BLK3D automatically and if yes, how?

Best regards


Dear Agisoft development team,

Especially in very narrow spaces spherical cameras having advantages over frame or fisheye cameras. I have used them to measure vertical channels in old buildings and other narrow spaces. But when you work with spherical cameras, MetaShape have to handle them as ideal cameras, where no camera calibration can be done. Because of their construction they have parallax and therefore issues in stitch zone, which downgrades accuracy and quality of results you can get with them. Some of them save the original fisheye photos as double (or more) fisheye photo. It is possible to split them to single fisheye photos and use them in MetaShape. But because of very small overlap and the missing possibilityto define constrains between a pair of fisheyes you can't align them properly without of the help of the stitched spherical photos.

So my feature request would be:
1) to split to, handle and calibrate left and right fisheyes (in case of double lens spherical cameras) as single cameras groups with their own calibration parameters. (I do that by hand now)
2) For alignment it is needed to define 6, before calculated, constrains, 3 rotations and 3 translations of the second lens in respect to the first lens. This way it is possible to align extracted fisheye photos alone without of the help of the imperfect spherical photos and their stitching errors.

Would be really nice, thanks!

Best regards


Feature Requests / Brightness enhancement for setting markers
« on: February 22, 2018, 01:39:38 PM »
In close range photogrammetry I have often difficulties to set markers in dark areas. It would be really helpful then to enhance the brightness and/or contrast temporary in the image like in other photogrammetry packages for better vizualisation of the area around the controlpoint.

Best regards


Face and Body Scanning / face scanning, new improvements
« on: January 25, 2016, 07:52:58 PM »
For those interested see the new article by Paul Graham, Graham Fyffe, Borom Tunwattanapong, Abhijeet Ghosh and Paul Debevec:
"Near-Instant Capture of High-Resolution Facial Geometry and Reflectance"



Trying a little experiment - making a set of scanned in historical photos available to see how good a model can be produced.

Details + dataset and provisional model available here:

Thank very very much for that Peter! I'm also very interested in how good a model can produced by the Agisoft team and other interested people. (With my standard version of PhotoScan I wasn't able to align all photos.)

Regards Karsten

General / See Agisoft at Intergeo 2014 Berlin
« on: October 06, 2014, 12:00:08 AM »
For those not aware off, you can see Agisoft at Intergeo Conference october 7th to 9th in Berlin. You can find Agisoft in hall 4.1 stand B4.001.


Feature Requests / easy support for pattern projection workflow
« on: September 28, 2014, 02:26:25 PM »
For smooth surfaces with little or none structure or texture, pattern projection would be a big improvement for Photoscan. A simplyfied workflow for reordering and renaming would be really helpful.

Because it surely isn't an easy task to group the images automatically, some simple ideas come to my mind to help Photoscan:

- the image content for an image with and without a projected pattern are significantly different and the grouping by image content is surely difficult. But this one with the lower feature count should be the image without pattern projected and this one with the higher feature count this one with projected pattern.

- When grouping by image content isn't possible (or is it?), what about using the time stamp, because images with and without pattern projection should be taken nearly at the same time. Could this be used to group images? Sometimes there are probably mistakes by the photographer, so that not every time exactly two images belong together and interaction by the user is necessary.

Thank you and regards


Feature Requests / separate pre-release installation
« on: September 28, 2014, 11:09:49 AM »
Separate installation of a pre-release version that one can try this pre-release and use the normal stable version for the daily work.

(With a little work it is possible also with the actual version, see here:, but could be simpler when it would be installed for example in a pre_release folder  :) )


General / Agisoft at 3D-ARCH 2013
« on: March 06, 2013, 08:16:24 PM »
Agisoft Photoscan has discussed by some speakers at 3D-ARCH 2013 - 3D Virtual Reconstruction and Visualization of Complex Architectures 25 – 26 February 2013, Trento, Italy.

For example:

You find a list of all presentations here:


Feature Requests / Working with mirrors
« on: January 04, 2013, 02:08:40 PM »
Sometimes you need to photograph something where your camera doesn't fit in. Then it would be very helpful to use an optical mirror. I tried this with a pot shooted from the mouth to the mirror inside the pot and the first step "Align Photos" works as expected. The positions of the cameras look as they were shot from around the pot. In the "Build Geometry" step the colored geometry is succesfully built (see first attachment). But the third step "Build Texture" failed, because Photoscan projects the texture from the camera position to the first face and not the right one (see second attachment).
I think it would be very helpful for many users, if Photoscan could produce the right texture also.

Thank you,


PS: If you need the data, I'd be glad to to send them to dropbox.

Feature Requests / Align image stack
« on: December 28, 2012, 12:40:59 PM »
For makro photogrammetry it would be nice, if the implemented keypoint generator or a similarly process could be used to align image stacks for expanding the depth of field.
Aligned image stacks are also useful for generating HDR images inside of Photscan and use them and they can be used for superresolution images ore noise reduction ore both.


Pages: [1]