With the impending CV implementation, the current "download every image we don't have" implementation is not feasible.
Both tj-uav/GroundStation (GS) and main.py (FS) need to be updated with such an improved implementation.
Specifics of a sample implementation:
The first request is made from GS->FS requesting for an update on images
FS responds with an update on both total images taken and any objects identified (with locations and the image number they're from)
GS does not pull every image from FS
GS either pulls every image that contains an object, or the sections of these images that contain the object
Exact implementation is specific on the implementation of tj-uav/CV.
GS displays downloaded images and their objects on the submissions page (Submissions Page Implementation GroundStation#149
Users can request to download specific images (or all images) with buttons
If only object-containing image sections are sent, users can request the full image
Related Issues:
With the impending CV implementation, the current "download every image we don't have" implementation is not feasible.
Both tj-uav/GroundStation (GS) and
main.py(FS) need to be updated with such an improved implementation.Specifics of a sample implementation:
Related Issues: