Issues generating DEM from Pleiades 1A stereo pairs using ASP

16 views
Skip to first unread message

Vico Alvarellos

unread,
Nov 7, 2025, 11:21:31 AMNov 7
to Ames Stereo Pipeline Support
Hi all, 

I'm working with ASP for a DEM generation using stereo pair Pleiades 1A imagery.
I'm quite new using these images and ASP and I'm struggling with some issues.

The images arrived separated in several files (R1C1, R1C2, etc). So I created a mosaic before running anything, as the tutorial suggests, so these were the following codes/steps:

1) gdalbuildvrt vrt77.tif 77201/IMG_PHR1A_P_001/IMG_PHR1A_P_202510161451124_SEN_7565677201-1_*R*C*.TIF

gdalbuildvrt vrt76.tif 76201/IMG_PHR1A_P_001/IMG_PHR1A_P_202510161450429_SEN_7565676201-1_R*C*.TIF

2) parallel_stereo -t pleiades --stereo-algorithm asp_mgm --subpixel-mode 9 77201/IMG_PHR1A_P_001/vrt77.tif 76201/IMG_PHR1A_P_001/vrt76.tif 77201/IMG_PHR1A_P_001/DIM_PHR1A_P_202510161451124_SEN_7565677201-1.XML 76201/IMG_PHR1A_P_001/DIM_PHR1A_P_202510161450429_SEN_7565676201-1.XML results/runtest

3) point2dem --t_srs epsg:32719 --tr 15 results/runtest-PC.tif

The output is a DEM full of holes.. Attached you wil find a picture of how it looks (img1). 

So I took a look into the PC file..:
  vico@srv-lamoge-01:~/Pleiades/1eraEntrega$ gdalinfo results/runtest-PC.tif | head -n 40 Driver: GTiff/GeoTIFF Files: results/runtest-PC.tif Size is 114151, 60756 Coordinate System is: GEOGCRS["WGS 84", ENSEMBLE["World Geodetic System 1984 ensemble", MEMBER["World Geodetic System 1984 (Transit)"], MEMBER["World Geodetic System 1984 (G730)"], MEMBER["World Geodetic System 1984 (G873)"], MEMBER["World Geodetic System 1984 (G1150)"], MEMBER["World Geodetic System 1984 (G1674)"], MEMBER["World Geodetic System 1984 (G1762)"], MEMBER["World Geodetic System 1984 (G2139)"], ELLIPSOID["WGS 84",6378137,298.257223563, LENGTHUNIT["metre",1]], ENSEMBLEACCURACY[2.0]], PRIMEM["Greenwich",0, ANGLEUNIT["degree",0.0174532925199433]], CS[ellipsoidal,2], AXIS["geodetic latitude (Lat)",north, ORDER[1], ANGLEUNIT["degree",0.0174532925199433]], AXIS["geodetic longitude (Lon)",east, ORDER[2], ANGLEUNIT["degree",0.0174532925199433]], USAGE[ SCOPE["Horizontal component of 3D system."], AREA["World."], BBOX[-90,-180,90,180]], ID["EPSG",4326]] Data axis to CRS axis mapping: 2,1 Origin = (0.500000000000000,-0.500000000000000) Pixel Size = (1.000000000000000,-1.000000000000000) Metadata: AREA_OR_POINT=Area POINT_OFFSET=1944708.7721009532 -5038455.0336882286 -3384728.7055590218 Image Structure Metadata: COMPRESSION=LZW INTERLEAVE=BAND  

I thought: as the PC file generated is huge, maybe it is better to process each R1C1, .. file. BUT, whenever I do that the DEM was correctly generated [:)] but a wrong georeferrenced output was obtained (close to the area, but not the exact one) and each R1C1, .. files were overlapping among each other. Attached you will find how it looks (img2).

Has anyone else experienced this? Do you know how can I fix it? It is better to work with the big area o process each tile and georreference it somehow.. 

Thanks a lot!!

Victoria 

Victoria
img1.jpgimg2.jpg

Oleg Alexandrov

unread,
Nov 7, 2025, 11:24:23 AMNov 7
to Vico Alvarellos, Ames Stereo Pipeline Support
It is suggested to try to mapproject the input images, and overlay them in a GIS viewer on top of the terrain before continuing. If that works, you can run stereo with mapprojected images. We have more details here: https://stereopipeline.readthedocs.io/en/latest/examples/pleiades.html#stereo-with-mapprojected-images

--
You received this message because you are subscribed to the Google Groups "Ames Stereo Pipeline Support" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ames-stereo-pipeline...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ames-stereo-pipeline-support/d69a4f6a-2def-493a-9cd0-7569740083e6n%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages