How to reduce OpenCV denoise processing time in Jetson Nano


I’m developing a product using Jetson Nano. (Env : python3 + OpenCV)

I have a question about OpenCV denoise processing.

It takes a long time to denoise with OpenCV.

  • cv2.fastNlMeansDenoisingColored()
  • Is there a way to reduce the processing time by using the CUDA or other functions to perform denoise processing?

    From system, you could try to enable the nvpmodel and jetson_clocks.

    Please refer to L4T documentation -> Power management.

    l4t document

    Thanks for your answer.

    I’ll check.
    and one more question.

    Does cv2.fastNlMeansDenoisingColored () function use GPU ?

    Please check it on openCV forum. Thanks.