AI Based Object Redaction - Examples - GStreamer Pipelines
| AI Based Object Redaction |
|---|
|
| Overview |
| Getting Started |
| GStreamer Plugin |
| Examples |
| Performance |
| Troubleshooting |
| FAQ |
| Contact Us |
GStreamer Pipelines
|
|
This section provide basic usage examples of pipelines to use the Object Redaction elements. The pipelines must be compound of one detection element followed by a redaction element.
Currently the plug-in only contains one detection element to detect faces and one redaction element to redact using a blurring algorithm.
Redact Video files
The following pipeline receives a input video file and outputs a MP4 video file with the redact faces.
gst-launch-1.0 filesrc location=input_video.mp4 ! qtdemux ! h264parse ! nvv4l2decoder ! nvvidconv ! queue ! rrfacedetector ! queue ! rrobjectredactor ! queue ! nvvidconv ! nvv4l2h264enc ! h264parse ! qtmux ! filesink location=output.mp4
The following pipeline receives a input video file and outputs a AVI video file with the redact faces.This video can be open with the Avidemux video player and editor.
gst-launch-1.0 filesrc location=input_video.mp4 ! qtdemux ! h264parse ! nvv4l2decoder ! nvvidconv ! queue ! rrfacedetector ! queue ! rrobjectredactor ! queue ! nvvidconv ! nvv4l2h264enc ! h264parse ! avimux ! filesink location=output.avi
Redact from camera input
The following pipeline receives an input video from a live camera and displays the redact video.
gst-launch-1.0 v4l2src ! nvvidconv ! queue ! rrfacedetector ! queue ! rrobjectredactor ! queue ! nvvidconv ! xvimagesink