Any plans to release a plugin example with actual image processing?

Hi, I know there is the gst-example plugin as an example how to write custom plugins. However, this plugin only reads from the image buffer and attaches metadata. I am trying to figure out how to write a plugin that actually manipulates the image (not assigning “transform_ip” but “transform” function), but have not even achieved passing the input buffer to the output buffer, I always get a segmentation fault somewhere down the pipeline when simply copying the input to the output buffer.

It would be really helpful to have a sample plugin that shows how to access the input buffer on the GPU and write it to the output buffer.

Thx in advance

bump

Hi skideep,

We’re discussing if can have a sample plugin provided internally, please stay tuned.

Here is gstnvsegvisual.cpp code

https://devtalk.nvidia.com/default/topic/1061638/deepstream-sdk/is-code-for-nvsegvisual-plugin-available-/post/5378098/#5378098