Forum

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 - mrc

Pages: [1]
1
General / Bad Texture after removing and regenerating texture
« on: February 19, 2020, 03:05:35 PM »
I have a number of models decimated to different face counts. When I remove a texture from one of these models and recreate it, frequently the text appears badly calculated, as shown in the screenshot below. The object is a bone, you can see the correct texture colour in some spots. The white lines are the edges of the texture segments in the UV map and should not be visible either.

I am removing the UVs before regenerating the textures, although I have moved from 4 x 4096 tiles to 1 x 8192 tiles.

Using version 1.6.1




2
General / Low number of depth maps produced resulting in missing surface
« on: December 18, 2019, 02:11:44 AM »
I have some scans I am reprocessing in version 1.5/1.6

When these were processed previously, I was getting 334 depth maps produced for 334 images (1.3). In 1.6 I am getting just 240 depth maps produced for the same set of photos, and this is leaving huge holes across my model. This happens whether I just do a "Depth Maps" mesh, or create a dense cloud for the model.

My parameters are Ultra-High, High, Mild with Interpolation turned on. All 334 of my images are aligned (Highest, no pairing, adaptive fitting).

I understand that some images are not used for depth mapping, but I would not expect the system to ignore images for which there is no overlap.

Edit: I ran a test without removing points using "Gradual Selection" in the sparse cloud phase and it seems to have generated Depth Maps for 325/334 images.
What are the criteria for whether Depth Maps are calculated from an image or not; so I can still optimise my alignment but don't remove so many points that the image is ignored?

3
This bug has been there forever, but it's driving me nuts because I have had to process around 10,000 photos in the last month or two.

When you select some photos in the bottom tabbed "Photos" window, the same photos are not always selected in the "Workspace" window. This means that if you filter some photos and select them and right click and select "Reset Camera Alignment", it's not the photos that you have selected that become reset - it's some other photos in the "Workspace" windows that are selected and reset.

If this could be fixed, it would be save me a lot of my coins going into the swear jar...

Cheers

4
General / Import Mask - smallest area to mask
« on: August 29, 2019, 11:17:55 AM »
I use masks heavily against a green-screen background. When I import the mask, I just import a single coloured image and let Metashape automatically mask my image.

I've noticed that small areas of "Green screen" do not get flagged as masks, though. Imagine a cup with a handle on a turntable. The gap in between the handle and the cup will get masked if the cup is perpendicular to the camera, but as the cup rotates and the empty area in between the handle and the cup gets smaller, there is a point where the software will no longer apply the mask even though a good deal of green-screen is still visible.

Is this a setting somewhere that I can adjust?

5
General / Low Resource Usage in new setup
« on: February 12, 2019, 09:43:00 AM »
I have had to partially rebuild a computer, specs are basically: Intel I7 8th Gen 4.2Ghz, 48GB DDR 4, nVidia GTX 1060 6G.

I processed one set of images, and the alignment stage was nice and quick, however I recently went to process a second set (375 x 24MP JPEG) and the alignment stage seems to have slowed. I opened resource monitor and also nVidia monitor, and the GPU is barely being used - sitting at 4%. CPU and RAM usage are also low - in fact the system doesn't appear to being fully utilised at all.

Anything else I can check?

6
Feature Requests / Export Units to PDF
« on: February 07, 2017, 12:48:50 PM »
Is it at all possible to add the exporting on units in PDF format?

Units are in metres in Photoscan but when exported to PDF they are in "model units" and I have to edit the PDF to change model units to metres.

Thank you.

7
General / Lumpy surface reconstruction
« on: October 03, 2016, 06:36:17 AM »
I am scanning in a number of bone specimens.

I have created 3D models for probably 300 specimens by now, mostly with good results. I am using a Nikon D5500 with a 60mm Macro Lens, a turntable and studio lighting. I usually shoot with f20 as this seems to be the sweet spot for the lens. Processing in high with gradual selection removing points above 0.3 Reprojection and 10 reconstruction uncertainty  in two iterations.

On a few specimens, I am getting really poor results. These usually occur when there is a smooth surface in the object, such as the shaft of a leg bone, or the surface of a skull. It is infinitely worse on smaller specimens ~2-4cm length This is especially the case for small objects. I usually take around 200 photos per specimen, and have tried many different approaches to remove these artefacts but nothing seems to work.

Below are two images, one of the photos and the same view of the dense point cloud reconstruction on High Alignment and High dense cloud. I have processed them in TIFF, DNG, JPG and using various depth filtering options to no avail.

How can I reduce this noise? Any idea what is causing this?

8
Feature Requests / Upload to Sketchfab - Private Flag
« on: July 16, 2016, 05:28:46 AM »
Is there a way to upload models direct to Sketchfab but make them private by default? I often upload models to sketchfab for showing progress to collaborators, but it seems any "followers" get a notification and direct link as soon as the model gets uploaded.

I think the Sketchfab API does have an option for this flag. It would be great if this was available as a feature. Thanks.

9
General / Order of Photos drastically affects reconstruction
« on: March 11, 2016, 02:10:15 AM »
I am using photogrammetry to capture a series of objects for scientific analysis, so my goal is to create a reliable method and be able to explain any anomalies.

I have a very consistent set up, with studio lighting, a decent SLR and lense, and a turntable. Objects (5-20cm across) are placed on the turntable and rotated with the camera on a tripod. I take between 120 and 300 images of each object depending on size and shape. Normally the reconstruction is good, but yesterday two reconstructions failed with either a large fuzzy sparse cloud, or a long line of points stretching off into infinity and all of the cameras clustered around a single point. (I am using masking).

After playing around with these for a few hours, I could not get them to align correctly. I tried different alignment options, different KPL values and accuracy settings - nothing helped. I reset all cameras and tried realigning them but also no good - this did stop the "shooting off" into space, but it kept resulting in a fuzzy point cloud, or with a second "ghost" point cloud of my object slightly offset from the other

I looked to see what the difference was between these scans and the 10 others that did work, and realised that in this scan, I had flipped the object once during the middle of the scan as I was unable to move the tripod at the time. My usual method is to take rings of photos moving the camera down at each ring, then flip the object and take more rings moving the camera up each time. These two I took 1 ring, flipped the object, took another ring, then moved the camera down and repeated.

With this in mind, I reset alignment on all cameras and then manually did them in the order as if I had taken them using the normal method. The alignment was perfect.

I repeated this several times on both scans and can confirm that this results in a near perfect reconstruction, where just processing them in order results in various bad reconstructions.

The order of the photos is obviously very important to the algorithm used for reconstruction. My questions are:

  • What is the optimum order for the photos to appear in to help the matching?
  • Does image rotation matter (I had a previous question on this), ie if I rotate the images with the object "flipped" does this help the reconstruction?
  • What is Photoscan using to order the photos for processing: filename, date and time, order of photos in the grid within the application?
  • Is there anything else I can do to "help" the reconstruction algorithm?

Sorry for the long winded post, thanks for your help.

10
General / Does image orientation matter?
« on: February 20, 2016, 09:40:23 AM »
Hi,

I am taking images of fossils for a research project.

As I need both sides of the fossil, I take a series of images of the fossil laying on its right hand side, then I flip it over and take a series of photos laying on its left. I then process all images in the same chunk - often this works. Sometimes, even though I use different backing material when I flip the fossil, Photoscan aligns the flipped fossil images upside down.

I am wondering if I rotate the "flipped" images 180 degrees, does this have any effect on the alignment?

Thanks

Pages: [1]