Because the business needs to be decoupled, I have to isolate the flow and detection here, so I can’t use the nvinfer component. I don’t need to process the monitoring image of each frame, I just need to extract the key frames. I tried uridecodebin->streammux->queue- >nvvideoconvert->nvjpegenc ->fakesink In this pipeline, the image is indeed compressed at the source of the stream, which is good, but later I found that there were many damaged images, so I referred to this post and yours The pipeline structure mentioned before uses nvstreamdemux, but the current status is that the stream is blocked
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Uridecodebin with filesink | 18 | 6048 | October 12, 2021 | |
| Jetson goes curling (or, simultaneously viewing multiple IP-cams) | 33 | 12284 | November 28, 2014 | |
| Seeking suggestions of faster way of saving images with jetson-inference camera processing | 17 | 5324 | October 18, 2021 | |
| Saving video streams | 2 | 1130 | October 12, 2021 | |
| How to avoid a delay in RTP streaming | 4 | 2176 | October 12, 2021 | |
| Deepstream+PyQt5+OpenCV - RTSP stream delay | 9 | 3153 | July 12, 2020 | |
| Gstdsexample turn tracked obj into base64 and Add to the DsMeta | 4 | 969 | October 12, 2021 | |
| Deepstream Python Bindings cannot access frame from deepstream-test-app3 | 17 | 3112 | October 12, 2021 | |
| Use Hardware rescaling of jetson nano in python codes | 11 | 900 | October 18, 2021 | |
| Deepstream python app for MJPG stream | 9 | 2497 | September 7, 2023 |