Forum Agisoft Metashape

Hello all, I am trying the new Classify Points... command on a small point cloud already ground classified. I am using Train Classifier option, selecting the current cloud as the training cloud. After some time, I get following error: Code: [Select] Error: Can'#039t open file: Le fichier spécifié est introuvable (2): D:/Guadalupe/Guadalupe_nadirs.files/1/0/dense_cloud.2/6_classes.bin Is this normal? See attached log file...What would be the normal production flow for such a case

This build does fix my issue with texturing. At some point, moving my project between versions had subtly broken the alignment so I need to redo the alignment from scratch to get a decent texture map that could be evaluated.

Hello, yes I was able to align, adjust and add the CPs. However to get good consistent results I had to uncheck all cameras and adjust with the 4 measured CPs. There is a constant shift in the X, Y of all camera coordinates, indicating that the airborne RTK process was not consistent with the RTK on your ground control. See https://1drv.ms/f/c/ed590dbfa76ca766/IgAK1E5WKIRdTJvDKnGrDkN1AS1AfeHnrG5DdRfgimEO-MU?e=oHBS0r

Hi Paul, Here is a link to a dataset: https://drive.google.com/drive/folders/1jxjTCAV90Yh-oaQyzQulUi9laADW4E33?usp=sharing Have fun! Thanks W

Yes wizza, if you send a link to the image dataset as well as markers with its coordinates then I could try and process on my side and give you my opinion.

Hi Paul, Some images attached which might help, would you like me to send you a link to a dataset? This data wasn'#039t really designed as a test project but for use in my usual photogrammetry workflow. Thanks W

What comes to mind is the Gcps were corrected using a different transformation. If you can share a few fotos with the markers, I could have a look at them....

Hi Paul, So the file you'#039ve sent matches with the original m4e coordinates, so I take it this means the transformation to osgb36 is correct? But then that still doesn'#039t explain the errors to the gcps. Anymore thoughts? Thanks again for your hel

Can you give a bit of insight into the reworked image matching? Has it improved overall alignment accuracy?

Can you give a bit of insight into the reworked image matching? Has it improved overall alignment accuracy?

Just to confirm, the Metastable double transfo checks with the official InquestII OS transformation within a mm in each axis!

Just to confirm, the Metastable double transfo checks with the official InquestII OS transformation within a mm in each axis!