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

Upgrade to OpenCV==4.10.0 #80

Draft
wants to merge 17 commits into
base: master
Choose a base branch
from
Draft

Upgrade to OpenCV==4.10.0 #80

wants to merge 17 commits into from

Conversation

BAILOOL
Copy link
Collaborator

@BAILOOL BAILOOL commented Aug 29, 2024

Closes #66

@BAILOOL BAILOOL added the enhancement New feature or request label Aug 29, 2024
@BAILOOL BAILOOL self-assigned this Aug 29, 2024
@BAILOOL BAILOOL changed the title Upgrade to Opencv==4.10.0 Upgrade to OpenCV==4.10.0 Aug 29, 2024
Copy link

github-actions bot commented Aug 29, 2024

LCOV of commit 0f7db5c during CI #172

Summary coverage rate:
  lines......: 83.0% (2938 of 3538 lines)
  functions..: 90.5% (248 of 274 functions)
  branches...: no data found

Files changed coverage rate: n/a

rameau-fr
rameau-fr previously approved these changes Aug 30, 2024
Copy link
Owner

@rameau-fr rameau-fr left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Great! Is the DockerHub updated too?

@BAILOOL BAILOOL dismissed rameau-fr’s stale review September 2, 2024 20:40

It's not ready yet due to failing unit tests related to handeye calibration

@BAILOOL
Copy link
Collaborator Author

BAILOOL commented Sep 3, 2024

@rameau-fr with OpenCV=4.10.0 we have some unit tests failing. It's probably related to hand-eye calibration not running.

The change to OpenCV was introduced in opencv/opencv#24897 where a part of the change was hard-coded filtering (with no possibility to change the thresholds).

if (Pgij_norm < 0.3 || Pcij_norm < 0.3 || Pgij_norm > 1.7 || Pcij_norm > 1.7) {
    continue;
}

@shreyas23
Copy link

Are there any expected quality upgrades from this PR?

@BAILOOL
Copy link
Collaborator Author

BAILOOL commented Sep 19, 2024

@shreyas23 TSAI hand-eye calibration method seems to be improved in the latest OpenCV. However, after the upgrade, our unit tests, where the accuracy is checked against ground truth (synthetic sequences), are failing.

Potentially, there is a bug in our code (#77, #56) and we will try to investigate it further when we have time.

Meanwhile, if you know of bugs or improvements to the codebase, feel free to list them. PR contributions are also welcomed.

@BAILOOL
Copy link
Collaborator Author

BAILOOL commented Oct 1, 2024

Seems that backwards compatibility can be achieved by setting board.setLegacyPattern(true); as mentioned in the issue opencv/opencv#23152

@BAILOOL
Copy link
Collaborator Author

BAILOOL commented Oct 1, 2024

Unit tests still do not pass. Potentially, there is more divergence between OpenCV versions despite setting board.setLegacyPattern(true) opencv/opencv#26126

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

Successfully merging this pull request may close these issues.

test not good in my dataset about fish images
4 participants