cellpose_omni.models

ARM	bool(x) -> bool
BD_MODEL_NAMES	Built-in mutable sequence.
C1_BD_MODELS	Built-in mutable sequence.
C1_MODELS	Built-in mutable sequence.
C2_BD_MODELS	Built-in mutable sequence.
C2_MODELS	Built-in mutable sequence.
C2_MODEL_NAMES	Built-in mutable sequence.
CP_MODELS	Built-in mutable sequence.
Cellpose([gpu, model_type, net_avg, device, ...])	main model which combines SizeModel and CellposeModel
CellposeModel([gpu, pretrained_model, ...])	param gpu whether or not to save model to GPU, will check if GPU available
MODEL_DIR	Path subclass for non-Windows systems.
MODEL_NAMES	Built-in mutable sequence.
MXNET_ENABLED	bool(x) -> bool
OMNI_INSTALLED	bool(x) -> bool
SizeModel(cp_model[, device, pretrained_size])	linear regression model for determining the size of objects in image used to rescale before input to cp_model uses styles from cp_model
cache_model_path(basename)
deprecation_warning_cellprob_dist_threshold(...)
model_path(model_type, model_index, use_torch)
models_logger	Instances of the Logger class represent a single logging channel.
size_model_path(model_type, use_torch)

cellpose_omni.io

check_dir(path)
get_image_files(folder[, mask_filter, ...])	find all images in a folder and if look_one_level_down all subfolders
get_label_files(img_names[, label_filter, ...])	Get the corresponding labels and flows for the given file images.
getname(path[, suffix])
imread(filename)
imsave(filename, arr)
imwrite(filename, arr, **kwargs)
load_links(filename)	Read a txt or csv file with label links. These should look like: 1,2 1,3 4,7 6,19 . . . Returns links as a set of tuples.
load_train_test_data(train_dir[, test_dir, ...])	Loads the training and optional test data for training runs.
logger_setup([verbose])
masks_flows_to_seg(images, masks, flows, ...)	save output of model eval to be loaded in GUI
outlines_to_text(base, outlines)
save_masks(images, masks, flows, file_names)	save masks + nicely plotted segmentation image to png and/or tiff
save_server([parent, filename])	Uploads a *_seg.npy file to the bucket.
save_to_png(images, masks, flows, file_names)	deprecated (runs io.save_masks with png=True)
write_links(savedir, basename, links)	Write label link file.

cellpose_omni.plot

disk(med, r, Ly, Lx)	returns pixels of disk with radius r and center med
dx_to_circ(dP[, transparency, mask, ...])	dP is 2 x Y x X => 'optic' flow representation
image_to_rgb(img0[, channels, channel_axis, ...])	image is 2 x Ly x Lx or Ly x Lx x 2 - change to RGB Ly x Lx x 3
interesting_patch(mask[, bsize])	get patch of size bsize x bsize with most masks
mask_overlay(img, masks[, colors, omni])	overlay masks on image (set image to grayscale)
mask_rgb(masks[, colors])	masks in random rgb colors
outline_view(img0, maski[, boundaries, ...])	Generates a red outline overlay onto image.
show_segmentation(fig, img, maski, flowi[, ...])	plot segmentation results (like on website)

cellpose_omni.metrics

aggregated_jaccard_index(masks_true, masks_pred)	AJI = intersection of all matched masks / union of all masks
average_precision(masks_true, masks_pred[, ...])	average precision estimation: AP = TP / (TP + FP + FN)
boundary_scores(masks_true, masks_pred, scales)	boundary precision / recall / Fscore
flow_error(maski, dP_net[, use_gpu, device])	error in flows from predicted masks vs flows predicted by network run on image
mask_ious(masks_true, masks_pred)	return best-matched masks

cellpose_omni.dynamics

compute_masks(dP, cellprob[, bd, p, inds, ...])	compute masks using dynamics from dP, cellprob, and boundary
follow_flows(dP[, mask, inds, niter, ...])	define pixels and run dynamics to recover masks in 2D
get_masks(p[, iscell, rpad, flows, ...])	create masks using pixel convergence after running dynamics
labels_to_flows(labels[, files, use_gpu, ...])	convert labels (list of masks or flows) to flows for training model
map_coordinates(I, yc, xc, Y)	bilinear interpolation of image 'I' in-place with ycoordinates yc and xcoordinates xc to Y
masks_to_flows(masks[, use_gpu, device])	convert masks to flows using diffusion from center pixel
masks_to_flows_cpu(masks[, device])	convert masks to flows using diffusion from center pixel Center of masks where diffusion starts is defined to be the closest pixel to the median of all pixels that is inside the mask.
masks_to_flows_gpu(masks[, device])	convert masks to flows using diffusion from center pixel Center of masks where diffusion starts is defined using COM :param masks: labelled masks 0=NO masks; 1,2,...=mask labels :type masks: int, 2D or 3D array
remove_bad_flow_masks(masks, flows[, ...])	remove masks which have inconsistent flows
steps2D(p, dP, inds, niter[, omni, calc_trace])	run dynamics of pixels to recover masks in 2D
steps2D_interp(p, dP, niter[, use_gpu, ...])
steps3D(p, dP, inds, niter)	run dynamics of pixels to recover masks in 3D

cellpose_omni.transforms

average_tiles(y, ysub, xsub, Ly, Lx)	average results of network over tiles
convert_image(x, channels[, channel_axis, ...])	return image with z first, channels last and normalized intensities
make_tiles(imgi[, bsize, augment, tile_overlap])	make tiles of image to run at test-time
move_axis(img[, m_axis, first])	move axis m_axis to first or last position
move_axis_new(a, axis, pos)	Move ndarray axis to new location, preserving order of other axes.
move_min_dim(img[, force])	move minimum dimension last as channels if < 10, or force==True
normalize99(Y[, lower, upper, omni])	normalize image so 0.0 is 0.01st percentile and 1.0 is 99.99th percentile
normalize_field(mu[, omni])
normalize_img(img[, axis, invert, omni])	normalize each channel of the image so that so that 0.0=1st percentile and 1.0=99th percentile of image intensities
original_random_rotate_and_resize(X[, Y, ...])	augmentation by random rotation and resizing X and Y are lists or arrays of length nimg, with dims channels x Ly x Lx (channels optional)
pad_image_ND(img0[, div, extra, dim])	pad image for test-time so that its dimensions are a multiple of 16 (2D or 3D)
random_rotate_and_resize(X[, Y, ...])	augmentation by random rotation and resizing
reshape(data[, channels, chan_first, ...])	reshape data using channels
reshape_and_normalize_data(train_data[, ...])	inputs converted to correct shapes for training and rescaled so that 0.0=1st percentile and 1.0=99th percentile of image intensities in each channel
reshape_train_test(train_data, train_labels, ...)	check sizes and reshape train and test data for training
resize_image(img0[, Ly, Lx, rsz, ...])	resize image for computing flows / unresize for computing dynamics
unaugment_tiles(y[, unet])	reverse test-time augmentations for averaging
update_axis(m_axis, to_squeeze, ndim)