Conversation
Signed-off-by: Sricharan Reddy Varra <sricharan.varra@biohub.org>
Signed-off-by: Sricharan Reddy Varra <sricharan.varra@biohub.org>
talonchandler
left a comment
talonchandler
left a comment
There was a problem hiding this comment.
Suggested next iteration:
- z tiling support
- benchmark with zyx tiles at zebrafish scale, with 3x3 through ~9x9 tiles...how long does i/o w/ and w/o blending take?
- a path to parallelization...it's okay if it's serial on a single job for now
If this is ready by next Wednesday, I think I can trial it on ~9 zebrafish tiles, maybe 512x512x256 per tile.
docs/examples/run_tile_single_fov.py
Outdated
| # -------------------- | ||
| # See how tiles are laid out with overlap. | ||
|
|
||
| slicer = Slicer(data, tile_size={"y": 32, "x": 64}, overlap={"y": 8, "x": 16}) |
There was a problem hiding this comment.
I'm unsure about if this Slicer is user facing or who should use it
There was a problem hiding this comment.
removed explicit mention of it in the examples.
docs/examples/run_tile_single_fov.py
Outdated
| result = map_tiles( | ||
| data, |
There was a problem hiding this comment.
Perhaps a better name? what did Jordao name his?
There was a problem hiding this comment.
Jordao called it apply_tiled, I've changed the name to apply_func_tiled.
| tile_size={"y": 32, "x": 64}, | ||
| overlap={"y": 8, "x": 16}, |
There was a problem hiding this comment.
What's the plan for "z" tiling? We'll need this for zebrafish.
There was a problem hiding this comment.
For z tiling not much really changes, I just needed to generalize the blending and the logic for creating composites in the overlap regions to support 3D.
Signed-off-by: Sricharan Reddy Varra <sricharan.varra@biohub.org>
Signed-off-by: Sricharan Reddy Varra <sricharan.varra@biohub.org>
2 participants
No description provided.