Closed unknownMarty closed 2 years ago
Hi,
I was looking at your jetson implementation. As far as I can see there is no DMA capability. Current setup rtsp source > gstreamer > decode > image buffer > convert to png > ultimate alpr
rtsp source > gstreamer > decode > image buffer > convert to png > ultimate alpr
references https://forums.developer.nvidia.com/t/gstreamer-nvmm-opencv-gpumat/52575/19 https://developer.ridgerun.com/wiki/index.php?title=GstInference/Example_pipelines/TX2
You don't need to convert the image to png, our sdk accepts a buffer as input. Check the signature of the process function at https://www.doubango.org/SDKs/anpr/docs/cpp-api.html.
Hi,
I was looking at your jetson implementation. As far as I can see there is no DMA capability. Current setup
rtsp source > gstreamer > decode > image buffer > convert to png > ultimate alpr
references https://forums.developer.nvidia.com/t/gstreamer-nvmm-opencv-gpumat/52575/19 https://developer.ridgerun.com/wiki/index.php?title=GstInference/Example_pipelines/TX2