we would like to process images in live, playback and for export as well.
Is it possible to retrieve the currently processed image, do some processing and forward the image back to the native Milestone pipeline for live, playback and export?
When you have video from a camera that is recorded, you cannot manipulate this video data that is recorded. XProtect being a VMS there is a promise that the data that originally came from the camera will not be altered.
So what is possible is the read this videodata, process it and then record it again as a new virtual camera.
If you look at the AI Bridge this is very much the pattern that is used.