Topics

1

General / cant open software

« on: February 22, 2021, 08:28:34 AM »

Hi:
Just want to open the psx file as normal, but then stuck in a white screen. Thought it should be a version upgrade problem, upgraded, still the same.

This is the first time I encounter the problem.

2

Python and Java API / How to cut the 3d model and tile models

« on: November 04, 2020, 10:23:55 AM »

Hi all:
I have lots of tilted aerial urban images, the 3d tile model of which is very bad on the four edges.
I want to cut them out by some means. Is there someway to do this? Such as defining an area of interest  or  resetting the bounding box through geographic coordinates.
much obliged 

3

General / What is the encoding format for .obj and other 3D models

« on: October 30, 2020, 12:18:25 PM »

Hi:
The 3d visulization software I use requires the 3d model encoding format.
THKs

4

Python and Java API / Orthomosaic translation

« on: September 01, 2020, 05:09:17 AM »

Hi
After orthomosaic computation, I want to export an orthomosaic with translated latitude and lontitude.
Theres nothing wrong with the orthomosaic computation or image position, its just I need a map translated in a certain way.
Somthing like

Code: [Select]

o = chunk.orthomosaic
o.left = o.left + 0.0001
o.right = o.left + 0.0001
o.top=o.left - 0.0001
o.bottom = o.bottom - 0.0001
But the objects in Metashape is not writable. Is there someway I can do this?

5

Python and Java API / exportRaster with jpeg compression

« on: June 14, 2020, 05:45:49 AM »

Hi, I want to export a quite large orthomosaic with the jpeg compression to reduce the size.
I used the following script, but it reported an error "cant write image".

Code: [Select]

import Metashape
chunk = Metashape.app.document.chunk
d_projection = Metashape.OrthoProjection()
d_projection.crs = Metashape.CoordinateSystem("EPSG::4326")
compression = Metashape.ImageCompression()
compression.tiff_compression = Metashape.ImageCompression.TiffCompressionJPEG
compression.jpeg_quality = 20
compression.tiff_big = True
compression.tiff_compression = True

chunk.exportRaster("C:/data/test.tif",
                    format = Metashape.RasterFormatTiles,
                    image_format = Metashape.ImageFormat.ImageFormatJPEG,
    image_compression = compression,
                    save_alpha = False,
                    white_background = False,
                    save_kml = False,
                    description =  " ",
                    projection = d_projection)


6

Python and Java API / Enable Metashape by floating license in Spyder

« on: October 23, 2019, 01:42:20 PM »

Halo:
I am trying to run the Metashape in the Spyder .
I have bunches of self-defined libraries, so it is not convenient to install them all again in the Metashape's python environment.

I am following the instructions in this link.
https://www.agisoft.com/forum/index.php?topic=10647.0

But the instructions only showed direct license key, while our department runs the floating license.
How can I activate Metashape in spyder in this case?

7

General / Why adjusted sensor location is zero after multilayer alignment

« on: September 26, 2019, 02:44:08 PM »

Halo:
I have 8 cameras organized as "multilayer" into the chunk with the GPS coordinates.
I hope to get the calibration values from the sparse alignment for further use to speed up the process.
But it seems the "adjusted location" field is always empty. What is the problem here, do I need to enable something?

8

General / Does an angular correction happen when computing the orthomosaic

« on: August 08, 2019, 12:29:57 PM »

Hallo everyone.
As the title, I am wondering if some kind of angular correction of the original image data happens when an orthomosaic (changed to different viewing angle) is computed?
If so, how can I disable it?
I would like to have only geometric projection of the pixels on the orthomosaic.

9

Python and Java API / how to distort the images again

« on: May 27, 2019, 03:30:22 PM »

Halo, I generated the undistorted images in Metashape because I wanted to apply some functions on these undistorted images.

Code: [Select]

for camera in cameras:
            ur=parent_path+camera.label+".tif"
            ref=camera.photo.image()
            refUndis=ref.undistort(camera.sensor.calibration,True,True)
            refUndis.save(ur)
But then how can I distort the fixed images again back? Becase I need to generate orthomosaic from them.

10

Python and Java API / Rotate every surface normal of a mesh model

« on: May 08, 2019, 08:38:05 AM »

Hi:
I have a lot of my aerial shots taken from a tilted angle, thus the default nadir orthomosaic cant represent my best view angle.
I recon that if every single vector in the mesh is rotated to a certain degree, then I can "nadir" project my images to this manipulated surface model.
Is there any python commands in metashape to achieve this?

11

Python and Java API / Define individual fixed parameters for calibration

« on: April 05, 2019, 01:46:44 PM »

Halo
I tried to import a part of calibration by a dict from laboratory precalibration. But when I double check it with graphic interface, I found all parameters are set. Then I tried remove sensor.fixed=True, no parameters are fixed. Isn't that sensor.fixed_params alone can determine the fixed parameters?

Code: [Select]

for s,sc in zip(doc.chunk.sensors, cali):
            calib=Metashape.Calibration()
            sc=cali[sc]
            intrinsic_p=sc["intrinsic_p"]
            calib.f=intrinsic_p["f"]
            calib.b1=intrinsic_p["b1"]
            calib.b2=intrinsic_p["b2"]
            calib.cx=intrinsic_p["cx"]
            calib.cy=intrinsic_p["cy"]
            distortion_p=sc["distortion_p"]
            calib.k1=distortion_p["k1"]
            calib.k2=distortion_p["k2"]
            calib.k3=distortion_p["k3"]
            calib.p1=distortion_p["p1"]
            calib.p2=distortion_p["p2"]     
            s.user_calib=calib
            s.fixed_params=["f","cx","cy","k1","k2","k3","p1","p2"]
            s.fixed=True
            doc.save()


12

General / What is the coordinate system for normal map

« on: April 04, 2019, 11:15:48 AM »

Hallo
If I understand correctly, a surface normal unit vector's (x,y,z) is encoded into RGB band by *255*2-1.

But what is this coordinate, is it just image coordinate itself or it is already world coordinate?

13

Python and Java API / Calibration with python

« on: February 14, 2019, 01:00:39 PM »

Hi:
I got the rotation angles and translation for the band alignment via a pre-calibration, and want to give these values to metashape with python. I am not very sure which should be the right place to give, theres a GPS/INS offset and slave offset.

My guts told me slave offset should be the one. But I couldn't find the object in python manual. While I did find the Metashape.Antenna for GPS/INS offset if I was not wrong. And I tested with that, where the result was very bad.

Any comment on that?

14

General / What is the rotation order and intial vector for camera pose

« on: December 06, 2018, 01:30:56 PM »

Hi:
I am trying to figure out an exact 3d vectors of the camera pose after photo alignment.
After exporting the georeference txt file to check the estimated yaw, pitch, roll values, I met a detail problem.
First, what is the order of the roation matrices multiplication.
And second, what is the initial vector that to be rotated? For instance, is that [0,0,-1] downwards or [1,0,0] xaxis positive.

15

General / Sparse vs Dense dem orthomosaic

« on: October 10, 2018, 12:01:44 PM »

Hi:
I tried to build orthomosaic image on dem generated from sparse cloud and dense cloud.
Sparse dem is faster, but I am not quite sure which is better in terms of quality, because they look quite same, except details. 
Can someone kindly provide some theoretical suggestions?