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.

Messages - Alexey Pasumansky

Pages: [1] 2 3 ... 616
General / Re: Why can't I see model after Photo Aignment?
« on: December 10, 2018, 12:55:20 AM »
Hello Ron,

Which coordinate system is set for the chunk in the Reference pane settings dialog?

Does it help, if select two not aligned cameras in the Reference pane, right-click on selection and choose Clear option to remove the coordinate information for them and then press Update button on the Reference pane toolbar?

General / Re: my system use only one (of two gpus)
« on: December 08, 2018, 07:37:43 PM »
i installed metashape 1.5.0 but i have not found  fine task level in advanced

Hello zagnimarco1,

It is in the bottom part of the dialog:

General / Re: Geotagging Images
« on: December 08, 2018, 04:06:20 PM »
Hello mwfoshee,

The CSV file that you are importing should contain the file extension for the images, not just camera labels as they appear in the Workspace pane, for example: DSC01013.JPG.

General / Re: Process four band images
« on: December 07, 2018, 03:30:34 PM »
Hello Murray,

I have changed the data type to U16 and corrected "doc" error in the code below (although that latter error shouldn't affect anything):
Code: [Select]
import PhotoScan
import os, glob
from PySide2 import QtWidgets

path_in ="Specify the path for input TIFFs:")
path_out ="Specify the path for output TIFFs:")

app = QtWidgets.QApplication.instance()
print("script started...")

image_list = [photo for photo in glob.iglob(path_in + "\\*.*", recursive = False) if (os.path.isfile(photo) and os.path.splitext(photo)[1][1:].lower() in ["tif"])]
input_photos = list()
for file in image_list:
image = PhotoScan.Image()
image =, datatype="U16", channels = "RGB ")
export_path = path_out + "//_" + os.path.basename(file)
print("Written image ", export_path)

chunk =
chunk.label = "Re-saved TIFFs"
chunk.addPhotos(input_photos) = chunk
print("script finished")

Python Scripting / Re: ANIMATION
« on: December 07, 2018, 01:13:19 PM »
Currently the .path file contains the following columns:
T, X, Y, Z, Qx, Qy, Qz, Qw

where T - is the time of the camera position (usually the values in this column go from 0 to the T_total - total duration of the track, with the regular step, i.e. frame rate),
X, Y, Z - coordinates of the camera location, Qx, Qy, Qz, Qw - quaternion that defines the orientation of the camera.

In case the project is georeferenced, the camera path should be defined in the LSE system (local coordinate system that is tangential to the ellipsoid).

Face and Body Scanning / Re: batch aligning chunks
« on: December 07, 2018, 01:02:19 PM »
Hello Brad,

Are you using "align chunks" in Batch process several times (for each chunk) or only once? And do you select the reference chunk there?

General / Re: Photoscan export creates pillar of points
« on: December 07, 2018, 11:25:46 AM »
Hello cwm,

Are you exporting points in WGS84 Lat/Long format?

I think that Cloud Compare doesn't support such representation and X, Y, Z units will be treated as equal in such case (for example, 54.56 degrees will be treated as 54.56 meters). Try to export the points either using Local Coordinates option or select a proper UTM projection in the Export Points dialog.

General / Re: How to change the Z axis over a non-referenced model?
« on: December 07, 2018, 11:23:37 AM »
Hello Ron,

You can use "Scale Object" tool ro apply arbitrary scale to the object or create a few scale-bars to apply real world scale to the object. After the scale is applied by any of these approaches you'll be able to generate DEM.

General / Re: my system use only one (of two gpus)
« on: December 07, 2018, 11:18:50 AM »
Hello zagnimarco1,

I don't think that Windows 10 Task Manager shows the GPU utilization in the default screen properly. Please try to go on the GPU page and switch one of the graphics there to Compute_0 or Compute_1. Or use dedicated utilities to monitor GPU activity, like GPU-Z or NVIDIA SMI, for example.
Additionally I recommend to disable CPU utilization in the GPU preferences tab when at least two GPUs are used.

To reduce the memory consumption, I can suggest to check version 1.5.0 pre-release and enable Fine-Level Task Distribution in the Advanced preferences tab. Also you can try new meshing method directly from the depth maps, skipping the dense point cloud.

General / Re: Export image quality
« on: December 06, 2018, 05:50:48 PM »
Hello ASW,

The new script version is the following:
Code: [Select]
import PhotoScan
chunk =
path ="Specify output path:)
file = open(path, "wt")
for camera in chunk.cameras:
   if "Image/Quality" in camera.meta.keys():
      file.write(camera.label + "\t" + camera.meta["Image/Quality"]+"\n")

Also you can just switch the Photos pane view to Details, select all lines, press CTRL+C and then copy the output (CTRL+V) to the external text editor.

General / Re: export Dense Point Cloud to xyz text file
« on: December 06, 2018, 03:06:47 PM »
Hello matteosin,

Currently the dense cloud cannot be transformed to the regular grid directly on export, but you can do the following:
generate DEM based on the dense cloud, then export DEM to ASCII xyz file with the custom resolution 1m/pix.

Shift fields are not intended for this. They should be only used, if you with to subtract the fixed value from the coordinates of each point in the cloud.

General / Re: How can I solve memory problem in "AgiSoft photo scan"?
« on: December 06, 2018, 03:04:33 PM »
Hello sherloc987,

Can you please specify the processing stage at which you do experience such issues and, if possible, provide the processing log from the Console pane and some general information about the project (number of images, image dimensions, processing settings used)?

General / Re: Photoscan vs 3DF Zephyr Lite - what am I doing wrong
« on: December 06, 2018, 03:03:26 PM »
Hello kezo,

It's a pre-release version:

General / Re: my system use only one (of two gpus)
« on: December 06, 2018, 02:35:43 PM »
Hello zagnimarco1,

Can you please provide the screenshot of the GPU preferences window from PhotoScan and also the processing log related to any GPU-supported task (image matching or depth maps generation, for example)?

General / Re: How to change the Z axis over a non-referenced model?
« on: December 06, 2018, 02:34:20 PM »
Hello Ron,

According to the screenshot your model is referenced (see [R] mark next to the chunk's label in the Workspace pane).

If you wish to discard the coordinate information for the camera locations in the Reference pane then please follow these steps:
- select all cameras in the Reference pane,
- right-click on selection and choose "Clear -> All" option from the context menu,
- open Reference pane settings dialog and switch the coordinate system to "Local Coordinates",
- in the Workspace pane right-click on the chunk's label and choose "Reset Transform" option from the context menu,
- use Rotate Object tool (located on toolbar) to adjust the model orientation according to the coordinate system axis in the bottom-right corner of the Model view.

Pages: [1] 2 3 ... 616