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

Improving Image Quality #10

Closed
atar13 opened this issue Oct 4, 2023 · 0 comments
Closed

Improving Image Quality #10

atar13 opened this issue Oct 4, 2023 · 0 comments
Assignees
Labels

Comments

@atar13
Copy link
Member

atar13 commented Oct 4, 2023

We want to take clear, high resolution images of ground targets. Here are a few constraints:

  • We must be flying at least 75 feet or higher (as per competition rules)
  • The plane will be moving in the ballpark of around 65 feet per second
  • There should be minimal blurring to give our computer vision algorithms the best chance at identifying the targets.

Here are some images taken from our LUCID Triton camera at a test flight last year: https://drive.google.com/drive/folders/1fiNQ4s1iYSufru__COxWTJQFbCT7pQyA?usp=drive_link

If you look at the images on the drive you'll see the images broken up into three main categories:

  1. The images have this wobble effect where some horizontal sections are slightly offset from other ones. We believe this is due to artifacts from a rolling shutter. The camera captures data in horizontal chunks and goes down in chunks. Here's on example of this artifact: https://drive.google.com/file/d/187FkuwvFchpmPcW6aeRHef0nSa3bCjFx/view?usp=drive_link
  2. Some images have some blurring that is a result of the movement of the plane while the image is being captured. These blurring artifacts are different than what we described with rolling shutter. Here's on image that has noticeable motion blur: https://drive.google.com/file/d/187FkuwvFchpmPcW6aeRHef0nSa3bCjFx/view?usp=drive_link
  3. A few of the images have targets inside them. In these cases, the targets are rarely exposed correctly. If the background is well lit, the targets are over-exposed (see the image on the right). The image on the bottom has targets that aren't as over-exposed, however the background is very poorly lit.
    sample targets
    This is a result of the camera's poor dynamic range. A small dynamic range means that the camera has difficulty capturing the differences between light signals that are very different from each other.

Here are some proposed steps we can take to remedy this problem:

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

No branches or pull requests

2 participants