Tiles

Fill in a module description here

 stitch (tiles_list:List[numpy.ndarray], x_list:List[int],
         y_list:List[int], iou_threshold:float)

Stitch tile masks into a global mask using IoU-based label matching.

	Type	Details
tiles_list	List	List of 2D mask arrays for each tile.
x_list	List	X-coordinates of tiles’ top-left corners.
y_list	List	Y-coordinates of tiles’ top-left corners.
iou_threshold	float	Minimum IoU threshold to match cell labels between overlapping tiles.
Returns	ndarray	Stitched mask with consistent global labels.

 save (tiles:List[numpy.ndarray], filename:str, x:List[int], y:List[int],
       output_dir:str)

Save a list of image tiles to disk with filenames encoding their positions.

	Type	Details
tiles	List	List of image tiles to save.
filename	str	Original filename to derive base name for tiles.
x	List	X-coordinates corresponding to each tile.
y	List	Y-coordinates corresponding to each tile.
output_dir	str	Directory to save the tile files.
Returns	None

 split (image:numpy.ndarray, tile_size:Tuple[int,int], overlap:int,
        tile_all:bool=True)

Split an image into tiles of specified size with a given overlap.

	Type	Default	Details
image	ndarray		Input image array (2D grayscale or 3D color).
tile_size	Tuple		Height and width of each tile, e.g., (128, 128).
overlap	int		Number of pixels each tile overlaps adjacent tiles.
tile_all	bool	True	If True, include partial edge tiles to cover the entire image. Default is True.
Returns	Tuple		Array of shape (N, tile_h, tile_w[, C]) containing all tiles.

 count (image:numpy.ndarray, t:int=1024, v:float=0.1)

Calculate the number of tiles in the vertical and horizontal directions.

	Type	Default	Details
image	ndarray		Input image array.
t	int	1024	Length (in pixels) of each square tile. Default is 1024.
v	float	0.1	Fractional overlap between tiles (e.g., 0.1 for 10%). Default is 0.1.
Returns	Tuple		Number of tiles along the vertical axis (height).