Easy Feature Tracking with Script TOP and OpenCV in TouchDesigner - Tutorial

Поділитися
Вставка
  • Опубліковано 4 лис 2024

КОМЕНТАРІ • 30

  • @harshinikarunaratne3928
    @harshinikarunaratne3928 2 роки тому +3

    Great video! How do I retrieve coordinates to use in a chop? (looking at eye tracking and need the coordinates of the gaze)

  • @LukeHildreth
    @LukeHildreth 4 роки тому +2

    Stoked to dive into this! Thanks for the technique ✌️

    • @TheInteractiveImmersiveHQ
      @TheInteractiveImmersiveHQ  4 роки тому

      So are we! Script TOP will unlock a lot of cool workflows we're excited to show off more :)

  • @jbach
    @jbach 4 роки тому +1

    Thanks Elburz

  • @kanelau7710
    @kanelau7710 3 роки тому

    That was very helpful, thx Elburz!

  • @Laserfxru
    @Laserfxru 3 роки тому +1

    Thank you!

  • @poststructural-ismlabel561
    @poststructural-ismlabel561 4 роки тому +1

    im trying to write the object detection using haar cascade and have a lot of problems, mainly for my lack of training... do you want to try out and make a tut please? literally there's no more video documentation about that new TOP script and opencv! thanks!

  • @zantaafio
    @zantaafio 4 роки тому +3

    sorry for offtopic but - am I the only one to wonder why the URL is "tutroals"? opencv-python-tutroals.readthedocs.io

  • @GlobeRCZ
    @GlobeRCZ Рік тому

    it works on MacOS? Should be great a tut to run OpenCV in TD on MacOS

    • @TheInteractiveImmersiveHQ
      @TheInteractiveImmersiveHQ  Рік тому

      It does, great suggestion! I've sent it over to the team for consideration as a future video topic :)

  • @avirads
    @avirads 3 роки тому

    hi , i am trying to convert a TD TOP to a base64 encoded image for showing it on a webpage...using the following code, but it does not encode properly..and the resultant image cannot be displayed.. can u please tell, what might be missing?
    import base64
    import cv2
    img = op('out1').numpyArray()
    encoded_img = cv2.imencode('.jpg',img)[1]
    print(base64.b64encode(encoded_img))

  • @dpredie
    @dpredie 3 роки тому +1

    Is it possible to use C++ instead of python for performance reasons?

    • @TheInteractiveImmersiveHQ
      @TheInteractiveImmersiveHQ  3 роки тому

      You could but that does add quite a bit over complexity both on the implementation in having to build C++ operators as well as having to know C++. If you're interested in that, there's a few good pages on the wiki about working with C++ operators: docs.derivative.ca/Write_a_CPlusPlus_Plugin

  • @LukeHildreth
    @LukeHildreth 4 роки тому +1

    How would I export the track points to use as a type of blob tracker?

    • @TheInteractiveImmersiveHQ
      @TheInteractiveImmersiveHQ  4 роки тому +1

      I had a lot of people ask me that so I'll make a quick video on that which should get posted on Monday :)

    • @LukeHildreth
      @LukeHildreth 4 роки тому

      @@TheInteractiveImmersiveHQ sounds good!

  • @marmaladereels
    @marmaladereels Рік тому

    mint toothpicks? living it large brutha, super deluxe

  • @kamiusu2360
    @kamiusu2360 Рік тому

    How to extract only points without video? And change their color?

    • @TheInteractiveImmersiveHQ
      @TheInteractiveImmersiveHQ  Рік тому +2

      Great question! To remove the video feed, you can add the line of code _img = cv2.subtract(img, op('null6').numpyArray(delayed=True))_ directly before the last line, _scriptOp.copyNumpyArray(img)_
      The color is set in the line _img[dst>0.01*dst.max()] = [0,0,255,255]_ , where [0,0,255,255] corresponds to the [Red, Green, Blue, Alpha] channels.
      Hope that helps!

    • @kamiusu2360
      @kamiusu2360 Рік тому

      @@TheInteractiveImmersiveHQ Thank you! It works great)

  • @lirawhite5799
    @lirawhite5799 Рік тому

    Thank you!