GRID_SPATIAL#
Python object: phenotypic.schema._grid_spatial.GRID_SPATIAL
Measure pixel-to-pixel distances to neighbors in adjacent grid cells.
For each detected colony, identify the nearest object in the left,
right, above, and below grid cells and report the minimum Euclidean
distance between their pixel masks. Distances are computed via a
per-section distance transform over a local window covering the target
cell and its immediate neighbors, so round colonies are not
over-estimated by their bounding boxes. Edge and corner colonies report
NaN for directions beyond the plate boundary.
Column label |
Description |
|---|---|
|
The object label of the left neighbor colony |
|
The minimum pixel-to-pixel distance to the left neighbor colony, computed via a Euclidean distance transform of object pixel masks |
|
The object label of the right neighbor colony |
|
The minimum pixel-to-pixel distance to the right neighbor colony, computed via a Euclidean distance transform of object pixel masks |
|
The object label of the above neighbor colony |
|
The minimum pixel-to-pixel distance to the above neighbor colony, computed via a Euclidean distance transform of object pixel masks |
|
The object label of the under neighbor colony |
|
The minimum pixel-to-pixel distance to the under neighbor colony, computed via a Euclidean distance transform of object pixel masks |