Getting started with NASA Ames Stereo Pipeline

This session can be adaptive to what folks want / need. In general, we will go over the basics of terrain model creation (what exactly it is that the software is doing), how to install ASP, and walk through a typical workflow for processing data. We’ll talk about data set preparation, parameter selection, and tips and tricks for making ASP work for you (rather than fighting against it).

Good references are our ASP paper and our GitHub repo. You can also find the PDF of the ASP Book as a separate download (if you only want to look at that without the whole package) under each release on the repo.

Ross’s slides can be found here: https://github.com/openplanetary/op-data-cafe/blob/master/epsc-dps2019/AmesStereoPipeline_RBeyer.pdf

Hello,
Currently i am working with ASP software to produce DEM from Pleiades imagery (installed on Linux server). Also i am interested on post processing DEM’s and orthoimage corrections like coregistration, filters and others.
To do this post-process step, i develop python scripts using Miniconda2&3 on windows 10 and i was wondering if its possible to install ASP into a conda environment? I am interesting to use use some functions of ASP into this conda environment, it is this possible?

I suppose that the answer is YES (i hope), but i don’t know where i have t pout the binaries of ASP and how to set up the PATH to use them.
If you have some help, i will appreciate it a lot.

Thank you in advance
Best regards

Hi Diego @cusicand. Welcome to the OpenPlanetary Forum and thanks for sharing your issue.
Unfortunately I don’t have have an answer/solution but @rbeyer, @michaelaye, @thare or @dpmayer might be able to help. I would also recommend that you cross post your question on https://astrodiscuss.usgs.gov.

This isn’t really an issue for Astrodiscuss, since Diego is talking about terrestrial Pleiades images, not images that are processed by ISIS at all.

ASP does not support installation via ‘conda install’ (maybe someday), but does have Linux and macOS binaries.

If you’re on a Windows machine, that means we don’t have binaries that will work for your system. However, some people have reported success using the Linux ASP binary installation in the Windows Subsystem for Linux (WSL), and other folks have used a Linux virtual machine environment to install ASP into.

If you’re talking about a conda environment on your Linux server, that’s a different story. If you are successfully using ASP on your Linux system regularly, once you conda activate an environment, it should be available via the same paths, etc.

It sounds like you’re talking about some Python that you have developed on Windows, and wanting to bring ASP to the Windows machine. However, Python is more portable than ASP, so it might be easier to develop your Python programs on your Windows machine, and then move them over to your Linux system where ASP works, and run them there.

Hi Ross @rbeyer, thank you for the detailed answer.

In fact, that is what i do, i have ASP installed on Windows Subsystem for Linux (WLS) in my personal machine and it works very good.
In all cases, i had curiosity about conda in order to have Python/ASP on the same environment on Windows 10.

Thank you very much for you help

Okay, in that case, the thing might be to establish a conda environment under WSL that supports your Python development (sure you may have to move things from your regular Windows conda environment to your WSL conda environment–not sure). Then inside of your WSL you should be able to conda activate and have access to your Python scripts and ASP (assuming conda works under WSL, maybe miniconda?).

Good luck!

There is no difference in the underlying tool for Anaconda/Miniconda. Both use the “conda” binary as the management tool, Anaconda just comes pre-installed with a huge meta-package pointing to approx 100 (data-)science Python packages. Here’s an apparent complete instruction set for Anaconda on WSL: https://gist.github.com/kauffmanes/5e74916617f9993bc3479f401dfec7da

Thanks a lot for the answer and your help @michaelaye and @rbeyer.
I’ll be back soon for sure :wink:.
Diego