Using FLIR Boson Camera input source with 'Y16 ' pixel format

• Hardware Platform (Jetson / GPU): Xavier AGX
• DeepStream Version: 6.0
• JetPack Version (valid for Jetson only): JP 4.6
• TensorRT Version:**

I am trying to run a FLIR boson using Deepstream. The camera is able to run with the ‘NV12’ format, however the camera applies gain in this format. The 'Y16 ’ format is the raw format.
When setting up config file with the resolution for 'Y16 ', the pipeline errors out. Any insight on how I can get deepstream to process 'Y16 ’ pixel format would be helpful.

Screenshot from 2022-01-24 15-35-38

*The top half is the source parameters in the config file. The bottom half is the error when deepstream-app runs with the config file

Please check if you can set to the format and run in gst-launch-1.0 like:

$ gst-launch-1.0 v4l2src device=/dev/video1 ! video/x-raw,format=GRAY16_LE,width=320,height=256,framerate=30/1 ! videoconvert ! video/x-raw,format=I420 ! nvvideoconvert ! nvoverlaysink sync=0

The possible formats are GRAY16_LE and GRAY16_BE. May give it a try and check if it works. Since DeepStream SDK is based on gstreamer, would need to find out a working pipeline. The format is special and it probably is not supported in v4l2src yet.

Hi DaneLLL,
Thank you for your help. The camera sends raw feed with the GRAY16_LE format! How can I set this up in the config file for deepstream-app?

Setting it in config file is not supported. Would need to customize the sample for a try. Please check create_camera_source_bin() in


And add foramt to cap1:

      caps1 = gst_caps_new_simple ("video/x-raw",
          "width", G_TYPE_INT, config->source_width, "height", G_TYPE_INT,
          config->source_height, "framerate", GST_TYPE_FRACTION,
          config->source_fps_n, config->source_fps_d, NULL);

If it works in gst-launch-1,.0 command, should also work by making the change. Please give it a try.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.