How can i draw skeleton from pose estimation result?

I am using Jetson Xavier and DeepStream.

I am using nvinfer for pedestrian detection, and its result boxes would be drawn well.

I want to infer pose estimation after pedestrian detection.

How can i infer the model and How can i draw skeleton result from pose estimation.

HI,

Here is a sample for your reference: https://github.com/tensorlayer/openpose-plus

In general, you will get the feature points of corresponding joints.
And you can show the skeleton by drawing the line of feature points.

Thanks.

In Deepstream plugin, how can i get the input buffer and draw and send to output buffer ?

Hi,

Please check this comment for the samples:
https://devtalk.nvidia.com/default/topic/1060956/deepstream-sdk/access-frame-pointer-in-deepstream-app/post/5375214/#5375214

Thanks.

how to draw lines in buffer?

When I change the input buffer value, it doesn’t work.

Hi,

You cannot change the INPUT buffer but output.
By the way, it’s recommended to use our OSD module to do the drawing.
https://docs.nvidia.com/metropolis/deepstream/4.0/dev-guide/index.html#page/DeepStream_Development_Guide%2Fdeepstream_app_config.3.2.html%23wwpID0EQHA

Thanks.