• Hardware Platform: Jetson Xavier NX
• DeepStream Version: 6.0.1
*• JetPack Version: 5.0.1 DP
• TensorRT Version: 8.4.0
Hi folks,
I have in my pipeline a nvvideoconvert element, I want to resize my stream using this element from 640x640 to 704x576 where 704 is the width and 576 is the height I read the document https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_plugin_gst-nvvideoconvert.html and look on dst-crop and src-crop property but I little confused.
Here is my code:
nvvidconv_postosd = Gst.ElementFactory.make("nvvideoconvert", "convertor_postosd")
nvvidconv_postosd.set_property("compute-hw", 2) # Using VIC
# Convert to OCU format 704X576
src_crop = "0:0:640:640" # Crop the entire input image (640x640)
dest_crop = "0:16:704:576" # Place the transformed image at (0, 16) in the output image (704x576)
nvvidconv_postosd.set_property("gpu-id", 0) # Set GPU ID if required
nvvidconv_postosd.set_property("src-crop", src_crop)
nvvidconv_postosd.set_property("dest-crop", dest_crop)
Unfortuntaly it’s look like the stream is 640x640 I will happy to modifcation and where I wrong.