Mismatch in elements in pool and in queue

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Hello,

I want to add pad probe to add my custom user meta before primary engine like sink pad of primary engine or source pad of streammux.

I do follow in example deepstream_user_metadata_app

But when using function api (gst_buffer_get_nvds_batch_meta). It return log then segment fault core dump

Mismatch in elements in pool and in queue

So I cannot add custom user meta to buffer. Any help! Thank you

1 Like
3

Got the same exact error while working with NvDsFrameMeta structures in NvDsBatchMeta beforehand.

4

This bug is my fault. I fixed this by do various things below and worked.

  • Check allocate memory and release data (in init metadata function, copy metadata function and release metadata function) of user metadata is right
  • With each value of each type should assign corresponding value (For me, my metadata is float but when assign 0 (int) value is not accepted, fix to 0.0.
  • Do not sure, but I add begin probe of sink pad (I not try add source pad again)