DeepStream-run inference on .mp4 and save the result as .mp4

Please provide complete information as applicable to your setup.

• Hardware Platform-GPU
• DeepStream Version-5.0
• TensorRT Version-7.0
• NVIDIA GPU Driver Version -440.33.01
• Issue Type- Video is saved without the inference Output in it.
pipeline - gst-launch-1.0 filesrc location=(pwd)/traffic_cam_4k.mp4 ! qtdemux ! h264parse ! nvv4l2decoder ! m.sink_0 nvstreammux name=m batch-size=1 width=1280 height=720 ! nvinfer config-file-path=(pwd)/config_infer_primary.txt ! nvvideoconvert ! nvdsosd ! nveglglessink
The above pipeline displays the correct output i.e. video with bounding boxes in it.
on the other hand if : gst-launch-1.0 filesrc location=(pwd)/traffic_cam_4k.mp4 ! qtdemux ! h264parse ! nvv4l2decoder ! m.sink_0 nvstreammux name=m batch-size=1 width=1280 height=720 ! nvinfer config-file-path=(pwd)/config_infer_primary.txt ! nvvideoconvert ! ‘video/x-raw(memory:NVMM),format=NV12’ ! nvv4l2h264enc ! h264parse ! qtmux ! filesink location=f1.mp4

This pipeline save the video named f1.mp4 BUT THERE IS NO BOUNDING BOXES IN THE VIDEO.

• How to reproduce the issue ?run the above mentioned pipeline on any .mp4 video

The bbox is drawn by nvdsosd plugin.Please add it to your pipeline if you want the bbox being drawn.

1 Like