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 - jinjamu

Pages: [1] 2
General / Black texture
« on: April 05, 2023, 01:22:11 PM »
Hi I am encountering a strange problem when texturing a mesh, the texture is turning out all black.  I have tried generating the mesh using interpolation off and on, same result.  The photos are of an underwater object, and because of the visibility the quality is quite low, however another model of an adjacent object taken on the same dive wiuth exactly the same camera etc turned out just fine.

Bug Reports / Incremental image alignment on 1.8.3
« on: June 23, 2022, 05:19:33 PM »
Are there any known issues with incremental image alignment on 1.8.3?
I have recently upgraded and I am taking a model built using 1.7.5, and I am trying to add a new set of images.  However the software is not doing any point detection and simply reporting that a number of images have not ben aligned.
Thanks in advance

Bug Reports / Blending Textures not using GPU
« on: May 19, 2022, 08:10:14 AM »
I am using the latest Metashape (1.8.3 build 14331) on a windows laptop with AMD Rhyzen CPU and a 3080 Laptop GPU. I have also upgraded to the latest Nvidia driver version 512.77.  I did this in an effort to sold this problem but the problem remains.
When running textures I am getting the following
Code: [Select]
Found 2 GPUs in 0.483 sec (CUDA: 0.224 sec, OpenCL: 0.259 sec)
Using device: NVIDIA GeForce RTX 3080 Laptop GPU, 48 compute units, free memory: 14911/16383 MB, compute capability 8.6
  driver/runtime CUDA: 11060/10010
  max work group size 1024
  max work item sizes [1024, 1024, 64]
No GPUs selected. Proceeding on CPU.
Done "load_model" in 0
rendering 4 pages
calculating mesh connectivity... done in 3.716 sec
I know for sure that I have not always had this problem, which must have been introduced at a specific point when upgrading, either the Metashape version or the Nvidia driver.

General / Error message "Incompatible keypoints"
« on: April 12, 2021, 06:53:44 PM »
I have been working on a rather large model of some underwater topography.  One of the steps of the process I am following involves adding cameras from a dive to a previous model (which is made up of all previous dives), and to then use a camera based alignment of the previous (accumulated) model to the new model (of the last dive) for the production of a rough model to help me to understand which areas are covered and which need further dives. 
When attempting this with my last data set I get an error "Incompatible keypoints".  The process I am following has worked fine for the first 7 dives so I cannot quite understand why I am encountering this problem now.  Any ideas would be appreciated.
Thanks and regards

Python and Java API / Script to disable every "nth"camera
« on: May 24, 2019, 05:02:56 PM »
I am working with some large datasets provided to me by somebody else.  On examining the images I realize that the amount of overlap is unnecessarily large,which will of course lead to unnecessarily long processing times.  Maybe there are other ways of tackling, however one way could be to run a script to disable every nth image (n could be maybe between 2 and 4) before running alignment.
I wish I was able to use python but I have not picked it up yet.  I was hoping that somebody can provide a script to do this?
Many thanks!

General / Manuals and installers for previous versions
« on: May 19, 2019, 09:40:30 PM »

Apologies if I am missing something, but are manuals and installers for older versions of Photoscan available for download.  I tend to use different versions for different parts of my workflow, and having ready access to a number of releases and their manuals would be very useful.


General / Bad Allocation error during mesh generation in 1.4.5
« on: April 13, 2019, 03:22:37 PM »

I am having this error coming up when running a mesh generation using the "consistent visibility mesh generation method" in 1.4.5, which I have had good success with in the past.

I am running on a highly specced VM on google cloud - 64 CPUs, 4 x Tesla P100, 256Gb RAM,  This is an expensive setup and it is pretty annoying that after running for around 20 hours I get this message!  Lots of money down the drain!

There was a solution posted here around increasing virtual memory which I have done, however before restarting the process only to have the same problem I was wondering if anyone had any further suggestions?

My model consists of around 2,400 24megapixel photographs.  Does anyone know the recommended memory required for such a size?

Many thanks!

General / Memory requirements for depth map model generation
« on: April 08, 2019, 08:02:01 PM »
Is there any rule of thumb to calculate memory requirements for the new "mesh generation based on depth maps" method?  Maybe based on number of photos, photo size, quality of mesh?

General / Strange behaviour when using manual markers for alignment
« on: March 24, 2019, 04:15:54 PM »

I'm working on quite a large model of an underwater shipwreck.  Because of the depth and limited time we have to shoot, I've had to combine photos from three separate dives into one model, and alignment has proved quite challenging.  The best result so far has come from combining all the photos (some 1.8k 24 or 36 MPixel photos) in one single alignment process (as opposed to working in chunks), but a number of photos did not align.  So I've tried using markers to force alignment between aligned and non-aligned photos, with mixed success. 

The strangest thing however, which I would like to ask about, happened following an alignment process which gave me a decent result but left a few photos out of alignment, when I selected the unaligned photos and did a "right-click...align selected cameras", I just got a "ball of cameras". 

I'm attaching screen grabs showing before and after.

Any ideas?


General / Masking transparent parts of images
« on: March 21, 2019, 09:27:22 AM »
I am trying to eliminate a large amount of unwanted background noise in a model by changing specific colours to transparent in a batch process, saving the images as PNGs, and redirecting Photoscan to use those images to texture the model.  The workflow seems to work well, only the "transparent" areas are being rendered as white, whereas I wish to eliminate the areas totally.  I am using Photoscan 1.4.5.  Does anyone have any suggestions or indeed maybe a different approach? Maybe there is a process where one can automatically mask out the transparent areas, but I dont know how to do this.

General / Info missing
« on: March 05, 2019, 09:58:43 AM »
Using Photoscan 1.4.5, and my info data (showing number of points in a model) has disappeared.  I've tried all sorts of things, even reinstalling, but for nothing.  The option is selected under the Model...Show Info setting.  This seemed to happen after I also installed Metashape on the same PC.
Any help would be appreciated

Bug Reports / Speed of 1.3.2 onwards vs 1.3.0
« on: December 30, 2017, 06:08:27 PM »
I'm currently doing a benchmarking exercise to compare speeds of different machines, including Azure based VMs, and in doing so I seem to be experiencing a massive difference in performance between v 1.3.0 versus v 1.3.2 & v1.3.4 (the versions I have tested) for the dense cloud build.

I have a fully aligned data set of 1509 photographs, which i am running through an Azure NV24 VM machine.  This is quite a powerful machine with 24 cores and 2 x Tesla M60 GPUs, see screen grab.

The dense cloud build comparative timings and resultant number of points is as follows:
1.3.0 - Depth Map 00:32:50, Dense Cloud 07:39:00, Points 56,043,481
1.3.4 - Depth Map 00:19:24, Dense Cloud 00:28:54, Points 26,606,884

So note the MASSIVE difference in time to build the dense cloud.

All processing parameters were of course identical.

Interesting also is that the depth map took substantially SHORTER time using 1.3.4, but also the number of resulting points is much less, around half, yet another strange thing. 

What I noticed during the processing of the depth maps on 1.3.0 was that the GPUs were maxing out at around 10-12% each...could explain the timing difference here.

But of course the most significant factor is the timing on the dense cloud part, and also the difference in point count.

Any ideas?  Is this a bug in 1.3.2 and later versions?




General / Older versions
« on: December 30, 2017, 10:37:09 AM »
Is there any way to download older versions of Photoscan?
The reason I am asking is that together with a colleague i am currently conducting some benchmarks on different machines,including Azure based VMs, and I am suspecting a significant performance drop in the dense cloud generation phase (at least with our with our test dataset), after version 1.3.0. 

General / Bringing a dense cloud back into photoscan
« on: December 02, 2017, 10:12:57 AM »
The workflow I am using involves doing the alignment with Photoscan, then exporting as a bundler.out format, and using other software to generate the dense cloud and to clean up that dense cloud.  This provides me with a really nice clean dense cloud.
My question is, is there a way to bring that clean dense cloud back into Photoscan in order to mesh and texture in Photoscan?
Many thanks!

General / Saving as high quality TIFF
« on: April 18, 2017, 10:48:56 PM »
I am using Photoscan Standard to document shipwrecks, and I would like to find a way of generating a high quality, large TIFF from the textured model.  There does not seem to be an easy way to do this (Export....TIFF), so I was wondering if anybody had any advice for me.

Pages: [1] 2