How to use ISP plugins for image processing with the gst-launch-1.0 tool

System version: jetpack 5.1.2
I now use the MIPI camera for image acquisition, the image display is not delicate enough, a little blurry, I want to adjust through ISP, see if the effect will be better, I use the following command to test is normal:

gst-launch-1.0 nvv4l2camerasrc device=/dev/video1  ! 'video/x-raw(memory:NVMM), width=(int)1920, height=(int)1080,interlace-mode=progressive, framerate=(fraction)30/1, format=(string)UYVY' ! nvvidconv  ! 'video/x-raw, format=(string)NV12' ! nv3dsink -e

But I want to test the ISP through the nvarguscamerasrc plugin is not normal, the video does not open, is it because the driver or device tree also needs to be configured? Or am I using the nvarguscamerasrc plugin incorrectly?

gst-launch-1.0 nvarguscamerasrc sensor-id=1 ! 'video/x-raw(memory:NVMM), \
     width=(int)1920, height=(int)1080, format=(string)NV12, \
     framerate=(fraction)30/1' ! nv3dsink -e

This is my device tree file, if I want to add additional link configuration if I want to go through the ISP? Or do you just need to use the plugin, is there a specific reference command? I tested a lot of instructions and it didn’t work, so I would like to ask for advice.
The main thing is to solve that the image is not clear enough, and the recognition effect is not very good when the object is recognized. thank you.
tegra194-p3737-0000-camera-tp2854vc-a00.dtsi (92.8 KB)
tegra194-p3737-tp285x-modules.dtsi (2.4 KB)

Dear NVIDIA,hi,adjust the video clarity captured by the MIPI Camera to make the recognized object better, how to process the image quality through ISP, or is there any other way to adjust? Thank you.

ISP pipeline don’t support YUV output format only support bayer raw output format.

Thanks