Search Unity

  1. Welcome to the Unity Forums! Please take the time to read our Code of Conduct to familiarize yourself with the forum rules and how to post constructively.
  2. Dismiss Notice

Question Frame improvement of OpenCV image processing (Hololens2)

Discussion in 'VR' started by Song_KCDR, Jun 1, 2021.

  1. Song_KCDR


    Feb 15, 2021
    I am trying to make a program to find objects using OpenCVForUnity on Hololens2.

    I tried running the example provided by OpenCVForUnity, and there was no problem in PC (800FPS) and smartphone (45FPS), but the frame rate was very low in Hololens2 (5FPS).
    Rendering the frame as-is has no problem with speed. Frames drop when image processing starts. (I tried with various types of samples)
    It was called in Update() every time because image processing was required in real time every frame.
    I made a simple thread and tested it, but there was some improvement, but it was not enough to be considered a solution. (10FPS)

    I also need to do image processing, I need spatial awareness, and I plan to do sound processing with a microphone. I don't know if it's possible to handle all of this on Hololens2. Do I need a network to do image processing on the server?

    Any advice on how I can increase the framerate when doing OpenCV on Hololens2?

    * Unity version uses 2019.4, and MRTK uses the latest version.
    Last edited: Jun 1, 2021