Showing results for 
Show  only  | Search instead for 
Did you mean: 
Registered: ‎03-31-2020

xlnxvideoscale and v4l2video0convert


I am working on the ZCU104 VCU ML platform (2019.2). My aim is try to use PL DDR for VCU decoding and map that decoded framebuffer to PS DDR for further processing (attached photo shows the Frame buffer read, VPSS and frame buffer write IP connected in the design). All what I want is to make VPSS to work (colorspace conversion and resize) the decoded frame in PL DDR and write back to PS DDR.

However, I am confusing the uses of xlnxvideoscale and v4l2video0convert (I could not get both the approaches work) as I am afraid that I am going to the wrong direction.


For the below command, would it work if I try to use PL DDR for VCU decoding and output to HDMI monitor? 
Because in PG252, the use case for PL DDR is using v4l2video0convert (in order to map to PS DDR for vcu encoding as decoded and encoder are using different DDRs)

gst-launch-1.0 filesrc location=demo_inputs/face_15fps.mp4 ! qtdemux ! h264parse ! omxh264dec internal-entropy-buffers=3 ! xlnxvideoscale ! \ video/x-raw, width=$WIDTH, height=$HEIGHT, format=BGR ! queue ! \ fpsdisplaysink video-sink="kmssink plane-id=37 bus-id="a2070000.v_mix" render-rectangle=\"<$(($WOFF + $((3*$WIDTH)))),$HOFF,$WIDTH,$HEIGHT>\"" text-overlay=false sync=false -v

However, from suggestions by , it seems that using v4l2video0convert would result in performance penalty, which is not expected. I think Chrisar's suggestion is correct for my case.

My question is: am I able to use xlnxvideoscale for VCU decoding and display on HDMI monitor utilizing PL DDR? Should Mem 2 mem VPSS Composite device is the what I should follow for this?( )

Screenshot from 2020-06-10 15-14-19.png






Tags (3)
0 Kudos
0 Replies