I’d like to use two Zed Boxes to sync four cameras (two on each box) into one API instance for unified skeleton tracking and depth images. Is this possible using only the 2 Zed Boxes, with one as master and one as slave, or does it require a 3rd machine to take input from both the Zed Boxes and process it separately as 4 networked cameras?
Yes, it is technically possible to use one of the ZED Box as “receiver” for the fusion module but I would not recommend it.
Indeed, this will add even more load on the box and can decrease the quality of detection (by reducing the FPS of the app), and degrading the fusion.
If possible, I would recommend using a 3rd machine to perform the fusion.