How do you select the frames before using Nuke Copycat?
I recently made an automatic frame recommendation system using OpenCV in Nuke.
You can easily download at the end of this page.
Nuke Copycat Auto frame? How does it work?
- It works by comparing the previous frame to the current frame and taking the sum of the differences
- The generated frames are in order of the largest sum.
- You can control how many frameholds are generated by adjusting the count value.
It is possible to determine which frames should be rotoscoped before training Copycat by using OpenCV.
To download and install the Nuke Copycat Auto frame recommendation
here is the pipeline:
- Python 3 and OpenCV must be installed before running this code.
- 1.Once you have downloaded this txt file, copy and paste it into your script editor
- 2. Select the plate and press the Auto Frame Detect button. (Currently only available for MOV files – will be updated)
- 3. OpenCV will calculate the frames that have significant differences between the previous frame and the current frame.
- 4. The resulting information shows which frames should be rotoscoped for better results.
It is still a work in progress to refine the details(baking this as a gizmo as well), but any ideas regarding this will be considered for further improvements!
- Made in Indie version, I will help if you message me with any challenges.
How to download Python3?
Please check this page and install Python on your machine.
How to install Open CV?
To install CV2, run the following command in a terminal window.
pip install opencv-python
please check this link for the demo video (version 1)
Download Auto Frame detection file
Thanks for reading, enjoy copy cat time!
*Other than for personal use, unauthorized commercial use is not permitted.
*If you have any questions, please contact firstname.lastname@example.org