Zero copy using CUDA allocated memory on Jetson TX2

[itaidagan]

Hello everyone,

I'm currently working on a plugin derived from VideoFlip that takes a frame,
copies the contents of the regular Mat to a GpuMat (using "upload"),
performs a cv::cuda::remap function and then copies the remapped GpuMat back
to the buffer frame (using "download").

I'd like to accelerate this process by avoiding the unnecessary copies. Is
there a way to tell gstreamer to allocate buffers using cudaMalloc in a way
that will allow me to implement this pipeline without any unnecessary
copies?

Here's the relevant code:

    // Upload Mat to GpuMat
    jmundistort->frame_gpu_mat.upload(jmundistort->frame_mat);

    // Remap
    cuda::remap(
            jmundistort->frame_gpu_mat,
            jmundistort->undistorted_frame_gpu_mat,
            jmundistort->dist_params.map1_gpu,
            jmundistort->dist_params.map2_gpu,
            INTER_CUBIC);

    // Download
    jmundistort->undistorted_frame_gpu_mat.download(jmundistort->frame_mat);




--
Sent from: http://gstreamer-devel.966125.n4.nabble.com/