Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Idea - Prioritize Building a Collection of Images from the (QHY183c/Meike 3.5mm/Acrylic Dome) #16

Open
danlohman opened this issue Jun 15, 2023 · 3 comments

Comments

@danlohman
Copy link
Collaborator

  • We should put a high priority on establishing an initial repository of video footage samples which can be used for development.
  • It's important that the image and video samples are produced by the same combination of optical components which have been specified for the station: the camera (QHY183c), fisheye lens (Meike 3.5mm) and protective acrylic dome.
  • Ideally images being produced by the actual station equipment we plan to use should be informing if not driving many of the decisions being made about the software.
  • For this reason, it would be great to start building out a set of video samples of the sky during various conditions (day, night, dusk, dawn, cloudy, clear, rainy, snowy, etc) so developers have them as a resource.
  • Sample images from the specific equipment combination that is planned for use also allows us to train ML models with the correct training data.
  • It also allows us to build and test the software in ways that model the specific of the distortion of the fisheye lens we are using and how it varies across manufacturing production runs (as long as we have the time and location and visible stars).
  • We will also start learning about any other unknown issues which are specific to the system as it is designed.
  • We can start with just a few images/videos of a selection of critical scenarios (a plane flying through the field of view during the day and at night, etc) then we can build from there.
  • The subsequent image processing pipeline and speed/performance requirements of each step needs is highly dependent on the initial images entering the pipeline.
  • What we see in the image data from the QHY183c/Meike 3.5mm Fisheye Lens and Acrylic dome will influence the requirements and the demands on the software that we are building and making decisions about now.
@danlohman danlohman added the Idea label Jun 15, 2023
@realsnick
Copy link
Contributor

although I think this is important, I would say having the alpha version complete is first. We don't want to rush things only to get a video out quick.

I think what @fcbarros is doing now with the QHY development app (to store images) is more then enough but further work should really be the recorder/playback itself

@realsnick
Copy link
Contributor

to summerize,

we need sample data for day/night with lat/long and timestamp.

@fcbarros i think for now your testing app should be enough . lat/long can be fixed and timestamp we already have as part of ROS

@realsnick
Copy link
Contributor

@danlohman i suggest to work with @ozzyjoe2874 to provide the data you need

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

2 participants