Help with Recording chunk video with filesrc with given frame number

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

• Hardware Platform NVIDIA 1650 4GB
• DeepStream Version 7.1

I am working with DeepStream 7.1 and want to record video when an object is detected (5 seconds before detect, and 5 seconds after detect). I explored the example at apps/sample_apps/deepstream-testsr, which records video using START_TIME and DURATION. However, I face the following challenges:

  • The example using RTSP stream, but when I use filesrc, so the recording length doesn’t match the expected duration of the video. Why?
  • Can I record multiple independent videos at the same time? (suppose frame_i I detect object, frame_i+1 I also detect. So it will record 2 videos)
  • I am considering using GstBuffer to manually save frames, but I am struggling with implementing.

Thanks

4
  1. AYK, a video file should start from I Frame. the app will dump no-I frames until find a I Frame.
  2. Currently, there is no support for overlapping smart record. please refer to the doc.
5

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

6

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