Motivation
The segmentation is currently decoupled from the acquisition and both takes time when they could be combined. The Raspberry Pi 5 can segment one image per second. Even if it's possible to create different segmenter, we can just refer here to the current legacy version with the flat calculation. Ideally we want to be able to run the live segmentation during the acquisition and keep the segmentation happening after the acquisition.
ToDo
Specifications
Live here means to segment images as they come from the acquisition.
acquisition n -> segmentation n
acquisition n+1 -> segmentation n+1
...
instead of
acquisition n -> acquisition n+1
then
segmentation n -> segmentation n+1
Motivation
The segmentation is currently decoupled from the acquisition and both takes time when they could be combined. The Raspberry Pi 5 can segment one image per second. Even if it's possible to create different segmenter, we can just refer here to the current legacy version with the flat calculation. Ideally we want to be able to run the live segmentation during the acquisition and keep the segmentation happening after the acquisition.
ToDo
Specifications
Live here means to segment images as they come from the acquisition.
acquisition n -> segmentation n
acquisition n+1 -> segmentation n+1
...
instead of
acquisition n -> acquisition n+1
then
segmentation n -> segmentation n+1