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 - stephan

Pages: 1 [2] 3
16
Feature Requests / Custom point classes?
« on: September 24, 2019, 12:29:09 PM »
Hi there!

I am curious: is this a feature that doesn't exist, or am I just reading the API docs wrong? 

I can't seem to find how to create additional point classes outside of: [Created, Unclassified, Ground, LowVegetation, MediumVegetation, HighVegetation, Building,
LowPoint, ModelKeyPoint, Water, Rail, RoadSurface, OverlapPoints, WireGuard, WireConductor, TransmissionTower, WireConnector, BridgeDeck, HighNoise, Car, Manmade]

Cheers,
Stephan

17
Python and Java API / import depth maps
« on: September 23, 2019, 05:42:19 PM »
Hi there!

I'm trying to use the addDepthMaps functionality to import depth maps that have been modified back to a PhotoScan project but I can't figure out how to use the finction using the API documentation.
I thought that I could go through the cameras one by one and use this function to attach an EXR file as depth map but this does not seem to be the case.
Does anyone have an example for me of how to do this?

Thanks!


18
Feature Requests / Export point clouds to PCD format
« on: August 19, 2019, 06:36:05 PM »
Hi all,

It would be useful to be able to export the dense and sparse PCs to PCD format directly from photoscan. PCD is becoming a standard used by many people so I think it would be helpful for a lot of people.

Cheers,

19
Python and Java API / Exporting point clouds to PCD format
« on: August 19, 2019, 06:34:15 PM »
Hi guys,

Does anyone here use PCD format? I am looking for an efficient way to export point clouds from Photoscan as PCD files, without calling a command-line tool from the python shell or doing any manual operations. Any tips?

Cheers,
Stephan

20
General / Metashape 1.5 build 7492 align issues
« on: January 11, 2019, 06:41:35 PM »
I just updated to the last version and it seems that it is broken... something went wrong between metashape preview build 7336 and current build 7492... I haven't changed any of my API codeand camera alibnments come out in a very inconsistent way: half of the time PhotoScan aligns only 20% of cameras and my project crashes. Does anyone else have issues ?

21
Python and Java API / importing pycuda to photoscan's python
« on: November 28, 2018, 01:44:51 PM »
Hi guys,

I am trying to setup a processing chain where I am using pycuda to do some image and point cloud manipulations. Right now I can use pycuda with the native python3 distribution in my OS and call this from photoscan as a shell script, however when I try to install pycuda inside the python version used by photoscan on Windows 10 I get compilation errors every time.
Has anyone had any luck setting this up? Is there something special that needs to be done to allow photoscan to access the VS c++ compiler from PATH?

Thanks!

22
Python and Java API / force refresh console ?
« on: August 24, 2018, 02:29:34 PM »
Hi guys,

Let's say I want to do this:

Code: [Select]
for i in range(10):
    print(str(i))
    time.sleep(0.5)


In the Photoscan Python console I get an output of all "print" functions once the complete loop is finished, ie after 10*0.5 seconds. Is there a way to force refresh the console after every iteration like a normal python console ?

Kind regards,

23
Python and Java API / Chance limit for depth map generation?
« on: July 26, 2018, 11:34:57 AM »
Hi guys,

From the photoscan manual I have found:

"Tie point limit parameter allows to optimize performance for the task and does not generally
effect the quality of the further model. Recommended value is 4000. Too high or too low tie point
limit value may cause some parts of the dense point cloud model to be missed. The reason is that
PhotoScan generates depth maps only for pairs of photos for which number of matching points is
above certain limit. This limit equals to 100 matching points, unless moved up by the figure "10%
of the maximum number of matching points between the photo in question and other photos,
only matching points corresponding to the area within the bounding box being considered."

Is there a way in the python API to set manually the value of number of Tie points to be taken into account for selection of image pairs that will be used to generate depth maps? Other than for performance, is there another reason to not use all image pairs ? I would be interested in seeing whether it is possible to improve quality by using more tie points but without missing any depth maps that would not be generated if there is less than 10% coverage between two images.

Thanks !

24
Python and Java API / export "real" depth maps?
« on: May 28, 2018, 05:57:35 PM »
Hi guys,
Is there a way to export the depth maps generated by Photoscan? I don't mean the "Depth images" created from each camera when there is a mesh in the chunk, but the actual depth maps generated by the buildDepthMaps step.
Thanks!

25
Python and Java API / Delete tie points outside of bounding box
« on: May 18, 2018, 06:08:15 PM »
Here is a puzzle I cannot figure out: I'd like to select all tie points that are a specific distance away from the closest edge of the bounding box, and to be able to set this distance as a function of the length along the X axis of the bounding box. Does anyone know how to make such a selection ?


26
General / Photoscan 1.3.1, GPU load reaching 20 to 30% max
« on: May 03, 2017, 07:16:11 PM »
Hi guys,
I'm running some tests on a Ryzen 1800X platform with dual Titan Xp graphics cards, 64Gb ram, 1200W PSU and I get about 20 to 30% load on each card maximum during dense point cloud generation. Any ideas what I could do to optimise this? Or is there just no way to max out 2 GPUs in this setup?
Thanks!

27
Python and Java API / select points by color
« on: September 22, 2016, 01:50:13 PM »
Hi guys,

I'm trying to test this function but no luck...


Code: [Select]

def main():
chunk = PhotoScan.app.document.chunk
#densepointcloud = PhotoScan.DenseCloud
PhotoScan.DenseCloud(0).selectPointsByColor(color=[255,255,255], tolerance=99, channels='RGB')
PhotoScan.DenseCloud(0).removeSelectedPoints()

main()

Any idea what I'm doing wrong?


28
Python and Java API / markers.position no longer working?
« on: July 01, 2016, 01:02:26 AM »
Hi,

This piece of code is causing trouble for some reason:

center1 = (chunk.markers[cam1].position + chunk.markers[cam2].position) / 2

I get this error:

TypeError: unsupported operand type(s) for +: 'NoneType' and 'NoneType'

This seems to be a new error as I didn't have it before updating, does anyone else have this?

29
Python and Java API / exportMasks?
« on: February 03, 2016, 09:06:12 PM »
Hi guys,

I can import masks using a python script but by the looks fo it there is no way to export them other than from the GUI... Am I overlooking something here or is the function just not implemented yet?

Cheers,
Stephan

30
Python and Java API / Define vector as Z axis
« on: January 06, 2016, 10:45:58 PM »
Hi guys,

I have a vector defined with the code below that I would like to make the models Z axis parallel to. Does anyone have any idea how this could be done?

Code: [Select]
    chunk.addMarker()
    marker1 = chunk.markers[0]
    chunk.markers[0].reference.location = PhotoScan.Vector([x,y,z])
   
    chunk.addMarker()
    marker2 = chunk.markers[1]
    chunk.markers[1].reference.location = PhotoScan.Vector([x1,y1,z1])
   
    updateregion.center = PhotoScan.Vector([x,y,z])

Pages: 1 [2] 3