Extracting Data from Featuring Tracking in TouchDesigner - Tutorial
ฝัง
- เผยแพร่เมื่อ 11 ต.ค. 2020
- Get access to 200+ hours of TouchDesigner video training, a private Facebook group where Elburz and Matthew Ragan answer all your questions, and twice-monthly group coaching/mastermind calls here: bit.ly/2zJ0lmt
A lot of folks are loving the new possibilities of the Script TOP, especially when it comes to implementing OpenCV workflows in realtime in TouchDesigner.
After our last video introducing the topic, we got a lot of questions on how to take the data from the feature detection and expose that in TouchDesigner for other workflows.
In this video, Elburz grabs another OpenCV example, ports it into TouchDesigner, and talks about two different methods of exposing that data.
And here it is! Dope, stoked for this direction.
:) Always glad we're going in the right direction!
Just listened to your episode on Notch's podcast! Helpful insights as always ✌️
Thanks! That was a fun one :)
Hey Elburz, I think I am running into a problem where the scriptTop doesn't cook when it's embedded inside a base/container. My hunch is that TD doesn't think the top needs to cook since it's not sending the top itself out. Right now my fix has been to send the scripTop out and put it in the flow so TD is pulling from it. Is this your experience as well? Thank you for the tutorials!!!
Thank you for this Video! This is good quick start guide.
It would be great if you can help with background segmentation. I'm struggeling with the BackgroundSubtractorMOG class in Td. :/
Let me give it a look. We have a video coming out on Monday that is using the QR code tracking with OpenCV and maybe that can help solve some of your issues (if it's the common CV_8U error I've seen folks having)
i have problems with Script top, error on line 17 which is:
img = op('null1').numpyArray(delayed=True) expected an indentent block
How about storing the data in an image at a specific rgb value so it is processed on the gpu?
You could also do that, since that's a bit more of an advanced workflow we kept this video geared towards getting the data in a format that is easiest for most people to use. You could probably do what you're mentioning directly in the Script TOP after you have the data you can overwrite pixels on your canvas instead of outputting a preview.
@@TheInteractiveImmersiveHQ great idea. I will tinker on this!
Oh blob tracking using this technique! I get some errors regarding some function with drawKeyPoints and 8bit/16bit errors
We have a new video coming up on Monday that talks about pixel formats in OpenCV that might help with the 8 bit / 16 bit errors. If that doesn't help, leave me another comment and I'll try to help :)
Can you turn auto-captions on? You have more than 1k users.
Let me take a look!
Thanks
@@fluttershift No problem!
You would need to ad concrete example of what you can do with this technique for people to understand, perhaps attaching a 2d party hat to your head with 30frames delay. Very interesting but not so usefull in real (only my opinion,anyone welcome to disagree) great lesson on scripting still