The goal is to not save recordings, hence save disk space, for events/alarms that our plugin-based AI integration analyzes and does not “validate”.
Can something be done with retention times?
The goal is to not save recordings, hence save disk space, for events/alarms that our plugin-based AI integration analyzes and does not “validate”.
Can something be done with retention times?
You can’t delete targeted video segments in Milestone. The best you can do is delete ALL video if you’re an administrator.
For this situation I’m thinking the best option is to use edge storage on the camera and have your ai integration make an edge retrieval request for video you deem important. Then the recording server will reach out to the camera and download the video and put it in the media database when you say so.
That requires cameras with edge storage and maintenance of edge storage SD cards though.
Another option is to take advantage of the prebuffer setting in Milestone and if the AI events are processed within 10-15 seconds you could have a rule that says “on AI event X, start recording 15 seconds before the event occurred and stop recording 15 seconds after the event stopped” or something like that. It’s not as precise but could be done without cameras with edge storage support.