Extract Analyzed Frame in nvdsanalytics-test

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) dGPU
• DeepStream Version 5.0
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only) 450.57

I have gone through the following links but could not start:



Can you please help on the following in the nvdsanalytics-test:

  1. Extracting the analyzed frames and save as jpg files.
  2. Enable Smart Records

Hi,
We would encourage users to do customization. Please make sure you can run default deepstream-nvdsanalytics-test first, and refer to guidance of enabling smart record:
https://docs.nvidia.com/metropolis/deepstream/dev-guide/index.html#page/DeepStream_Development_Guide/deepstream_smart_video_record.html

And the links of demonstrating saving to JPEGs.

I have been able to run the deepstream-nvdsanalytics-test and do some customisation. I have read through the Smart Record and DS-Example documentation and the source code but its difficult to start. Do you have some more hints, if yes, it would be really helpful in getting me going.

Hi,
For enabling smart record, currently we have document and the sample is undergoing. You would need to check the document and do the implementation.

For saving to JPG file, the samples in the links calls cv::imwrite(). In 5.0 GA, we have new functions in

deepstream-5.0\sources\includes\nvds_obj_encode.h

Samples:

deepstream-5.0\sources\apps\sample_apps\deepstream-image-meta-test
deepstream-5.0\sources\apps\sample_apps\deepstream-transfer-learning-app

Since samples are there and public, would encourage you take a look and do customization.

The header deepstream-5.0\sources\includes\nvds_obj_encode.h contains definitions for only crop object and save. How do we achieve saving the complete image with bounding boxes?

Hi,

We have given suggestion in


Need to use cv::imwrite() in this case. Please take a look.