MouseLand / cellpose / 149

pending completion

Build # 149

Build Type

Pull #254

travis-ci

Committed by

web-flow

Commit Message

<a href="https://github.com/MouseLand/cellpose/commit/<a class=hub.com/MouseLand/cellpose/commit/49bbe25a8f8f0feefd43d54329c3f4f7e8e58227">49bbe25a8<a href="https://github.com/MouseLand/cellpose/commit/49bbe25a8f8f0feefd43d54329c3f4f7e8e58227">">Merge </a><a class="double-link" href="https://github.com/MouseLand/cellpose/commit/<a class="double-link" href="https://github.com/MouseLand/cellpose/commit/f88b15a16a9453cbfe7c5d16c5e0f2b6c82a2be5">f88b15a16</a>">f88b15a16</a><a href="https://github.com/MouseLand/cellpose/commit/49bbe25a8f8f0feefd43d54329c3f4f7e8e58227"> into 012dc1b9b">012dc1b9b</a>

Pull Request Pull Request #254: Enable look_one_level_down during training as well

Run Details

3 of 3 new or added lines in 1 file covered. (100.0%)

1397 of 2579 relevant lines covered (54.17%)

1.63 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

44.49

/cellpose/io.py

import os, datetime, gc, warnings, glob
from natsort import natsorted
import numpy as np
import cv2
import tifffile
import logging, pathlib, sys

from . import utils, plot, transforms

try:
    from PyQt5 import QtGui, QtCore, Qt, QtWidgets
    GUI = True
except:
    GUI = False

try:
    import matplotlib.pyplot as plt
    MATPLOTLIB = True
except:
    MATPLOTLIB = False
    
try:
    from google.cloud import storage
    SERVER_UPLOAD = True
except:
    SERVER_UPLOAD = False

io_logger = logging.getLogger(__name__)
io_logger.setLevel(logging.DEBUG)

def logger_setup():
    cp_dir = pathlib.Path.home().joinpath('.cellpose')
    cp_dir.mkdir(exist_ok=True)
    log_file = cp_dir.joinpath('run.log')
    try:
        log_file.unlink()
    except:
        print('creating new log file')
    logging.basicConfig(
                    level=logging.INFO,
                    format="%(asctime)s [%(levelname)s] %(message)s",
                    handlers=[
                        logging.FileHandler(log_file),
                        logging.StreamHandler(sys.stdout)
                    ]
                )
    logger = logging.getLogger(__name__)
    logger.info(f'WRITING LOG OUTPUT TO {log_file}')
    #logger.handlers[1].stream = sys.stdout

    return logger, log_file

def outlines_to_text(base, outlines):
    with open(base + '_cp_outlines.txt', 'w') as f:
        for o in outlines:
            xy = list(o.flatten())
            xy_str = ','.join(map(str, xy))
            f.write(xy_str)
            f.write('\n')

def imread(filename):
    ext = os.path.splitext(filename)[-1]
    if ext== '.tif' or ext=='tiff':
        img = tifffile.imread(filename)
        return img
    else:
        try:
            img = cv2.imread(filename, -1)#cv2.LOAD_IMAGE_ANYDEPTH)
            if img.ndim > 2:
                img = img[..., [2,1,0]]
            return img
        except Exception as e:
            io_logger.critical('ERROR: could not read file, %s'%e)
            return None

def imsave(filename, arr):
    ext = os.path.splitext(filename)[-1]
    if ext== '.tif' or ext=='tiff':
        tifffile.imsave(filename, arr)
    else:
        cv2.imwrite(filename, arr)

def get_image_files(folder, mask_filter, imf=None, look_one_level_down=False):
    """ find all images in a folder and if look_one_level_down all subfolders """
    mask_filters = ['_cp_masks', '_cp_output', '_flows', mask_filter]
    image_names = []
    if imf is None:
        imf = ''
    
    folders = []
    if look_one_level_down:
        folders = natsorted(glob.glob(os.path.join(folder, "*/")))
    folders.append(folder)

    for folder in folders:
        image_names.extend(glob.glob(folder + '/*%s.png'%imf))
        image_names.extend(glob.glob(folder + '/*%s.jpg'%imf))
        image_names.extend(glob.glob(folder + '/*%s.jpeg'%imf))
        image_names.extend(glob.glob(folder + '/*%s.tif'%imf))
        image_names.extend(glob.glob(folder + '/*%s.tiff'%imf))
    image_names = natsorted(image_names)
    imn = []
    for im in image_names:
        imfile = os.path.splitext(im)[0]
        igood = all([(len(imfile) > len(mask_filter) and imfile[-len(mask_filter):] != mask_filter) or len(imfile) < len(mask_filter) 
                        for mask_filter in mask_filters])
        if len(imf)>0:
            igood &= imfile[-len(imf):]==imf
        if igood:
            imn.append(im)
    image_names = imn

    if len(image_names)==0:
        raise ValueError('ERROR: no images in --dir folder')
    
    return image_names
        
def get_label_files(image_names, mask_filter, imf=None):
    nimg = len(image_names)
    label_names0 = [os.path.splitext(image_names[n])[0] for n in range(nimg)]

    if imf is not None and len(imf) > 0:
        label_names = [label_names0[n][:-len(imf)] for n in range(nimg)]
    else:
        label_names = label_names0
        
    # check for flows
    if os.path.exists(label_names0[0] + '_flows.tif'):
        flow_names = [label_names0[n] + '_flows.tif' for n in range(nimg)]
    else:
        flow_names = [label_names[n] + '_flows.tif' for n in range(nimg)]
    if not all([os.path.exists(flow) for flow in flow_names]):
        flow_names = None
    
    # check for masks
    if os.path.exists(label_names[0] + mask_filter + '.tif'):
        label_names = [label_names[n] + mask_filter + '.tif' for n in range(nimg)]
    elif os.path.exists(label_names[0] + mask_filter + '.png'):
        label_names = [label_names[n] + mask_filter + '.png' for n in range(nimg)]
    else:
        raise ValueError('labels not provided with correct --mask_filter')
    if not all([os.path.exists(label) for label in label_names]):
        raise ValueError('labels not provided for all images in train and/or test set')

    return label_names, flow_names


def load_train_test_data(train_dir, test_dir=None, image_filter=None, mask_filter='_masks', unet=False, look_one_level_down=True):
    image_names = get_image_files(train_dir, mask_filter, image_filter, look_one_level_down)
    nimg = len(image_names)
    images = [imread(image_names[n]) for n in range(nimg)]

    # training data
    label_names, flow_names = get_label_files(image_names, mask_filter, imf=image_filter)
    nimg = len(image_names)
    labels = [imread(label_names[n]) for n in range(nimg)]
    if flow_names is not None and not unet:
        for n in range(nimg):
            flows = imread(flow_names[n])
            if flows.shape[0]<4:
                labels[n] = np.concatenate((labels[n][np.newaxis,:,:], flows), axis=0) 
            else:
                labels[n] = flows
            
    # testing data
    test_images, test_labels, image_names_test = None, None, None
    if test_dir is not None:
        image_names_test = get_image_files(test_dir, mask_filter, image_filter, look_one_level_down)
        label_names_test, flow_names_test = get_label_files(image_names_test, mask_filter, imf=image_filter)
        nimg = len(image_names_test)
        test_images = [imread(image_names_test[n]) for n in range(nimg)]
        test_labels = [imread(label_names_test[n]) for n in range(nimg)]
        if flow_names_test is not None and not unet:
            for n in range(nimg):
                flows = imread(flow_names_test[n])
                if flows.shape[0]<4:
                    test_labels[n] = np.concatenate((test_labels[n][np.newaxis,:,:], flows), axis=0) 
                else:
                    test_labels[n] = flows
    return images, labels, image_names, test_images, test_labels, image_names_test



def masks_flows_to_seg(images, masks, flows, diams, file_names, channels=None):
    """ save output of model eval to be loaded in GUI 

    can be list output (run on multiple images) or single output (run on single image)

    saved to file_names[k]+'_seg.npy'
    
    Parameters
    -------------

    images: (list of) 2D or 3D arrays
        images input into cellpose

    masks: (list of) 2D arrays, int
        masks output from Cellpose.eval, where 0=NO masks; 1,2,...=mask labels

    flows: (list of) list of ND arrays 
        flows output from Cellpose.eval

    diams: float array
        diameters used to run Cellpose

    file_names: (list of) str
        names of files of images

    channels: list of int (optional, default None)
        channels used to run Cellpose    
    
    """
    
    if channels is None:
        channels = [0,0]
    
    if isinstance(masks, list):
        for k, [image, mask, flow, diam, file_name] in enumerate(zip(images, masks, flows, diams, file_names)):
            channels_img = channels
            if channels_img is not None and len(channels) > 2:
                channels_img = channels[k]
            masks_flows_to_seg(image, mask, flow, diam, file_name, channels_img)
        return

    if len(channels)==1:
        channels = channels[0]

    flowi = []
    if flows[0].ndim==3:
        Ly, Lx = masks.shape[-2:]
        flowi.append(cv2.resize(flows[0], (Lx, Ly), interpolation=cv2.INTER_NEAREST)[np.newaxis,...])
    else:
        flowi.append(flows[0])
    if flows[0].ndim==3:
        cellprob = (np.clip(transforms.normalize99(flows[2]),0,1) * 255).astype(np.uint8)
        cellprob = cv2.resize(cellprob, (Lx, Ly), interpolation=cv2.INTER_NEAREST)
        flowi.append(cellprob[np.newaxis,...])
        flowi.append(np.zeros(flows[0].shape, dtype=np.uint8))
        flowi[-1] = flowi[-1][np.newaxis,...]
    else:
        flowi.append((np.clip(transforms.normalize99(flows[2]),0,1) * 255).astype(np.uint8))
        flowi.append((flows[1][0]/10 * 127 + 127).astype(np.uint8))
    if len(flows)>2:
        flowi.append(flows[3])
        flowi.append(np.concatenate((flows[1], flows[2][np.newaxis,...]), axis=0))
    outlines = masks * utils.masks_to_outlines(masks)
    base = os.path.splitext(file_names)[0]
    if masks.ndim==3:
        np.save(base+ '_seg.npy',
                    {'outlines': outlines.astype(np.uint16) if outlines.max()<2**16-1 else outlines.astype(np.uint32),
                        'masks': masks.astype(np.uint16) if outlines.max()<2**16-1 else masks.astype(np.uint32),
                        'chan_choose': channels,
                        'img': images,
                        'ismanual': np.zeros(masks.max(), np.bool),
                        'filename': file_names,
                        'flows': flowi,
                        'est_diam': diams})
    else:
        if images.shape[0]<8:
            np.transpose(images, (1,2,0))
        np.save(base+ '_seg.npy',
                    {'img': images,
                        'outlines': outlines.astype(np.uint16) if outlines.max()<2**16-1 else outlines.astype(np.uint32),
                     'masks': masks.astype(np.uint16) if masks.max()<2**16-1 else masks.astype(np.uint32),
                     'chan_choose': channels,
                     'ismanual': np.zeros(masks.max(), np.bool),
                     'filename': file_names,
                     'flows': flowi,
                     'est_diam': diams})    

def save_to_png(images, masks, flows, file_names):
    """ deprecated (runs io.save_masks with png=True) 
    
        does not work for 3D images
    
    """
    save_masks(images, masks, flows, file_names, png=True)

def save_masks(images, masks, flows, file_names, 
               png=True, tif=False, outlines=False):
    """ save masks + nicely plotted segmentation image to png and/or tiff

    if png, masks[k] for images[k] are saved to file_names[k]+'_cp_masks.png'

    if tif, masks[k] for images[k] are saved to file_names[k]+'_cp_masks.tif'

    if png and matplotlib installed, full segmentation figure is saved to file_names[k]+'_cp.png'

    only tif option works for 3D data
    
    Parameters
    -------------

    images: (list of) 2D, 3D or 4D arrays
        images input into cellpose

    masks: (list of) 2D arrays, int
        masks output from Cellpose.eval, where 0=NO masks; 1,2,...=mask labels

    flows: (list of) list of ND arrays 
        flows output from Cellpose.eval

    file_names: (list of) str
        names of files of images
    
    """
    
    if isinstance(masks, list):
        for image, mask, flow, file_name in zip(images, masks, flows, file_names):
            save_masks(image, mask, flow, file_name, png=png, tif=tif, outlines=outlines)
        return
    
    if masks.ndim > 2 and not tif:
        raise ValueError('cannot save 3D outputs as PNG, use tif option instead')
    base = os.path.splitext(file_names)[0]
    exts = []
    if masks.ndim > 2 or masks.max()>2**16-1:
        png = False
        tif = True
    if png:    
        exts.append('.png')
    if tif:
        exts.append('.tif')

    # convert to uint16 if possible so can save as PNG if needed
    masks = masks.astype(np.uint16) if masks.max()<2**16-1 else masks.astype(np.uint32)
    
    # save masks
    with warnings.catch_warnings():
        warnings.simplefilter("ignore")
        for ext in exts:
            imsave(base + '_cp_masks' + ext, masks)

    if png and MATPLOTLIB and not min(images.shape) > 3:
        img = images.copy()
        if img.ndim<3:
            img = img[:,:,np.newaxis]
        elif img.shape[0]<8:
            np.transpose(img, (1,2,0))
        
        fig = plt.figure(figsize=(12,3))
        # can save images (set save_dir=None if not)
        plot.show_segmentation(fig, img, masks, flows[0])
        fig.savefig(base + '_cp_output.png', dpi=300)
        plt.close(fig)

    if masks.ndim < 3 and outlines: 
        mask_outlines = utils.outlines_list(masks)
        outlines_to_text(base, mask_outlines)

def save_server(parent=None, filename=None):
    """ Uploads a *_seg.npy file to the bucket.
    
    Parameters
    ----------------

    parent: PyQt.MainWindow (optional, default None)
        GUI window to grab file info from

    filename: str (optional, default None)
        if no GUI, send this file to server

    """
    if parent is not None:
        q = QtGui.QMessageBox.question(
                                    parent,
                                    "Send to server",
                                    "Are you sure? Only send complete and fully manually segmented data.\n (do not send partially automated segmentations)",
                                    QtGui.QMessageBox.Yes | QtGui.QMessageBox.No
                                  )
        if q != QtGui.QMessageBox.Yes:
            return
        else:
            filename = parent.filename

    if filename is not None:
        os.environ['GOOGLE_APPLICATION_CREDENTIALS'] = os.path.join(os.path.dirname(os.path.realpath(__file__)),
                                                        'key/cellpose-data-writer.json')
        bucket_name = 'cellpose_data'
        base = os.path.splitext(filename)[0]
        source_file_name = base + '_seg.npy'
        io_logger.info(f'sending {source_file_name} to server')
        time = datetime.datetime.now().strftime("%Y_%m_%d_%H_%M_%S.%f")
        filestring = time + '.npy'
        io_logger.info(f'name on server: {filestring}')
        destination_blob_name = filestring
        storage_client = storage.Client()
        bucket = storage_client.bucket(bucket_name)
        blob = bucket.blob(destination_blob_name)

        blob.upload_from_filename(source_file_name)

        io_logger.info(
            "File {} uploaded to {}.".format(
                source_file_name, destination_blob_name
            )
        )

1	import os, datetime, gc, warnings, glob	3✔
2	from natsort import natsorted	3✔
3	import numpy as np	3✔
4	import cv2	3✔
5	import tifffile	3✔
6	import logging, pathlib, sys	3✔
7
8	from . import utils, plot, transforms	3✔
9
10	try:	3✔
11	from PyQt5 import QtGui, QtCore, Qt, QtWidgets	3✔
12	GUI = True	3✔
13	except:	×
14	GUI = False	×
15
16	try:	3✔
17	import matplotlib.pyplot as plt	3✔
18	MATPLOTLIB = True	3✔
19	except:	×
20	MATPLOTLIB = False	×
21
22	try:	3✔
23	from google.cloud import storage	3✔
24	SERVER_UPLOAD = True	3✔
25	except:	×
26	SERVER_UPLOAD = False	×
27
28	io_logger = logging.getLogger(__name__)	3✔
29	io_logger.setLevel(logging.DEBUG)	3✔
30
31	def logger_setup():	3✔
32	cp_dir = pathlib.Path.home().joinpath('.cellpose')	3✔
33	cp_dir.mkdir(exist_ok=True)	3✔
34	log_file = cp_dir.joinpath('run.log')	3✔
35	try:	3✔
36	log_file.unlink()	3✔
37	except:	3✔
38	print('creating new log file')	3✔
39	logging.basicConfig(	3✔
40	level=logging.INFO,
41	format="%(asctime)s [%(levelname)s] %(message)s",
42	handlers=[
43	logging.FileHandler(log_file),
44	logging.StreamHandler(sys.stdout)
45	]
46	)
47	logger = logging.getLogger(__name__)	3✔
48	logger.info(f'WRITING LOG OUTPUT TO {log_file}')	3✔
49	#logger.handlers[1].stream = sys.stdout
50
51	return logger, log_file	3✔
52
53	def outlines_to_text(base, outlines):	3✔
54	with open(base + '_cp_outlines.txt', 'w') as f:	×
55	for o in outlines:	×
56	xy = list(o.flatten())	×
57	xy_str = ','.join(map(str, xy))	×
58	f.write(xy_str)	×
59	f.write('\n')	×
60
61	def imread(filename):	3✔
62	ext = os.path.splitext(filename)[-1]	3✔
63	if ext== '.tif' or ext=='tiff':	3✔
64	img = tifffile.imread(filename)	3✔
65	return img	3✔
66	else:
67	try:	3✔
68	img = cv2.imread(filename, -1)#cv2.LOAD_IMAGE_ANYDEPTH)	3✔
69	if img.ndim > 2:	3✔
70	img = img[..., [2,1,0]]	3✔
71	return img	3✔
72	except Exception as e:	×
73	io_logger.critical('ERROR: could not read file, %s'%e)	×
74	return None	×
75
76	def imsave(filename, arr):	3✔
77	ext = os.path.splitext(filename)[-1]	3✔
78	if ext== '.tif' or ext=='tiff':	3✔
79	tifffile.imsave(filename, arr)	3✔
80	else:
81	cv2.imwrite(filename, arr)	3✔
82
83	def get_image_files(folder, mask_filter, imf=None, look_one_level_down=False):	3✔
84	""" find all images in a folder and if look_one_level_down all subfolders """
85	mask_filters = ['_cp_masks', '_cp_output', '_flows', mask_filter]	3✔
86	image_names = []	3✔
87	if imf is None:	3✔
88	imf = ''	3✔
89
90	folders = []	3✔
91	if look_one_level_down:	3✔
92	folders = natsorted(glob.glob(os.path.join(folder, "*/")))	3✔
93	folders.append(folder)	3✔
94
95	for folder in folders:	3✔
96	image_names.extend(glob.glob(folder + '/*%s.png'%imf))	3✔
97	image_names.extend(glob.glob(folder + '/*%s.jpg'%imf))	3✔
98	image_names.extend(glob.glob(folder + '/*%s.jpeg'%imf))	3✔
99	image_names.extend(glob.glob(folder + '/*%s.tif'%imf))	3✔
100	image_names.extend(glob.glob(folder + '/*%s.tiff'%imf))	3✔
101	image_names = natsorted(image_names)	3✔
102	imn = []	3✔
103	for im in image_names:	3✔
104	imfile = os.path.splitext(im)[0]	3✔
105	igood = all([(len(imfile) > len(mask_filter) and imfile[-len(mask_filter):] != mask_filter) or len(imfile) < len(mask_filter)	3✔
106	for mask_filter in mask_filters])
107	if len(imf)>0:	3✔
108	igood &= imfile[-len(imf):]==imf	×
109	if igood:	3✔
110	imn.append(im)	3✔
111	image_names = imn	3✔
112
113	if len(image_names)==0:	3✔
114	raise ValueError('ERROR: no images in --dir folder')	×
115
116	return image_names	3✔
117
118	def get_label_files(image_names, mask_filter, imf=None):	3✔
119	nimg = len(image_names)	3✔
120	label_names0 = [os.path.splitext(image_names[n])[0] for n in range(nimg)]	3✔
121
122	if imf is not None and len(imf) > 0:	3✔
123	label_names = [label_names0[n][:-len(imf)] for n in range(nimg)]	×
124	else:
125	label_names = label_names0	3✔
126
127	# check for flows
128	if os.path.exists(label_names0[0] + '_flows.tif'):	3✔
129	flow_names = [label_names0[n] + '_flows.tif' for n in range(nimg)]	×
130	else:
131	flow_names = [label_names[n] + '_flows.tif' for n in range(nimg)]	3✔
132	if not all([os.path.exists(flow) for flow in flow_names]):	3✔
133	flow_names = None	3✔
134
135	# check for masks
136	if os.path.exists(label_names[0] + mask_filter + '.tif'):	3✔
137	label_names = [label_names[n] + mask_filter + '.tif' for n in range(nimg)]	×
138	elif os.path.exists(label_names[0] + mask_filter + '.png'):	3✔
139	label_names = [label_names[n] + mask_filter + '.png' for n in range(nimg)]	3✔
140	else:
141	raise ValueError('labels not provided with correct --mask_filter')	×
142	if not all([os.path.exists(label) for label in label_names]):	3✔
143	raise ValueError('labels not provided for all images in train and/or test set')	×
144
145	return label_names, flow_names	3✔
146
147
148	def load_train_test_data(train_dir, test_dir=None, image_filter=None, mask_filter='_masks', unet=False, look_one_level_down=True):	3✔
149	image_names = get_image_files(train_dir, mask_filter, image_filter, look_one_level_down)	3✔
150	nimg = len(image_names)	3✔
151	images = [imread(image_names[n]) for n in range(nimg)]	3✔
152
153	# training data
154	label_names, flow_names = get_label_files(image_names, mask_filter, imf=image_filter)	3✔
155	nimg = len(image_names)	3✔
156	labels = [imread(label_names[n]) for n in range(nimg)]	3✔
157	if flow_names is not None and not unet:	3✔
158	for n in range(nimg):	×
159	flows = imread(flow_names[n])	×
160	if flows.shape[0]<4:	×
161	labels[n] = np.concatenate((labels[n][np.newaxis,:,:], flows), axis=0)	×
162	else:
163	labels[n] = flows	×
164
165	# testing data
166	test_images, test_labels, image_names_test = None, None, None	3✔
167	if test_dir is not None:	3✔
168	image_names_test = get_image_files(test_dir, mask_filter, image_filter, look_one_level_down)	×
169	label_names_test, flow_names_test = get_label_files(image_names_test, mask_filter, imf=image_filter)	×
170	nimg = len(image_names_test)	×
171	test_images = [imread(image_names_test[n]) for n in range(nimg)]	×
172	test_labels = [imread(label_names_test[n]) for n in range(nimg)]	×
173	if flow_names_test is not None and not unet:	×
174	for n in range(nimg):	×
175	flows = imread(flow_names_test[n])	×
176	if flows.shape[0]<4:	×
177	test_labels[n] = np.concatenate((test_labels[n][np.newaxis,:,:], flows), axis=0)	×
178	else:
179	test_labels[n] = flows	×
180	return images, labels, image_names, test_images, test_labels, image_names_test	3✔
181
182
183
184	def masks_flows_to_seg(images, masks, flows, diams, file_names, channels=None):	3✔
185	""" save output of model eval to be loaded in GUI
186
187	can be list output (run on multiple images) or single output (run on single image)
188
189	saved to file_names[k]+'_seg.npy'
190
191	Parameters
192	-------------
193
194	images: (list of) 2D or 3D arrays
195	images input into cellpose
196
197	masks: (list of) 2D arrays, int
198	masks output from Cellpose.eval, where 0=NO masks; 1,2,...=mask labels
199
200	flows: (list of) list of ND arrays
201	flows output from Cellpose.eval
202
203	diams: float array
204	diameters used to run Cellpose
205
206	file_names: (list of) str
207	names of files of images
208
209	channels: list of int (optional, default None)
210	channels used to run Cellpose
211
212	"""
213
214	if channels is None:	×
215	channels = [0,0]	×
216
217	if isinstance(masks, list):	×
218	for k, [image, mask, flow, diam, file_name] in enumerate(zip(images, masks, flows, diams, file_names)):	×
219	channels_img = channels	×
220	if channels_img is not None and len(channels) > 2:	×
221	channels_img = channels[k]	×
222	masks_flows_to_seg(image, mask, flow, diam, file_name, channels_img)	×
223	return	×
224
225	if len(channels)==1:	×
226	channels = channels[0]	×
227
228	flowi = []	×
229	if flows[0].ndim==3:	×
230	Ly, Lx = masks.shape[-2:]	×
231	flowi.append(cv2.resize(flows[0], (Lx, Ly), interpolation=cv2.INTER_NEAREST)[np.newaxis,...])	×
232	else:
233	flowi.append(flows[0])	×
234	if flows[0].ndim==3:	×
235	cellprob = (np.clip(transforms.normalize99(flows[2]),0,1) * 255).astype(np.uint8)	×
236	cellprob = cv2.resize(cellprob, (Lx, Ly), interpolation=cv2.INTER_NEAREST)	×
237	flowi.append(cellprob[np.newaxis,...])	×
238	flowi.append(np.zeros(flows[0].shape, dtype=np.uint8))	×
239	flowi[-1] = flowi[-1][np.newaxis,...]	×
240	else:
241	flowi.append((np.clip(transforms.normalize99(flows[2]),0,1) * 255).astype(np.uint8))	×
242	flowi.append((flows[1][0]/10 * 127 + 127).astype(np.uint8))	×
243	if len(flows)>2:	×
244	flowi.append(flows[3])	×
245	flowi.append(np.concatenate((flows[1], flows[2][np.newaxis,...]), axis=0))	×
246	outlines = masks * utils.masks_to_outlines(masks)	×
247	base = os.path.splitext(file_names)[0]	×
248	if masks.ndim==3:	×
249	np.save(base+ '_seg.npy',	×
250	{'outlines': outlines.astype(np.uint16) if outlines.max()<2**16-1 else outlines.astype(np.uint32),
251	'masks': masks.astype(np.uint16) if outlines.max()<2**16-1 else masks.astype(np.uint32),
252	'chan_choose': channels,
253	'img': images,
254	'ismanual': np.zeros(masks.max(), np.bool),
255	'filename': file_names,
256	'flows': flowi,
257	'est_diam': diams})
258	else:
259	if images.shape[0]<8:	×
260	np.transpose(images, (1,2,0))	×
261	np.save(base+ '_seg.npy',	×
262	{'img': images,
263	'outlines': outlines.astype(np.uint16) if outlines.max()<2**16-1 else outlines.astype(np.uint32),
264	'masks': masks.astype(np.uint16) if masks.max()<2**16-1 else masks.astype(np.uint32),
265	'chan_choose': channels,
266	'ismanual': np.zeros(masks.max(), np.bool),
267	'filename': file_names,
268	'flows': flowi,
269	'est_diam': diams})
270
271	def save_to_png(images, masks, flows, file_names):	3✔
272	""" deprecated (runs io.save_masks with png=True)
273
274	does not work for 3D images
275
276	"""
277	save_masks(images, masks, flows, file_names, png=True)	×
278
279	def save_masks(images, masks, flows, file_names,	3✔
280	png=True, tif=False, outlines=False):
281	""" save masks + nicely plotted segmentation image to png and/or tiff
282
283	if png, masks[k] for images[k] are saved to file_names[k]+'_cp_masks.png'
284
285	if tif, masks[k] for images[k] are saved to file_names[k]+'_cp_masks.tif'
286
287	if png and matplotlib installed, full segmentation figure is saved to file_names[k]+'_cp.png'
288
289	only tif option works for 3D data
290
291	Parameters
292	-------------
293
294	images: (list of) 2D, 3D or 4D arrays
295	images input into cellpose
296
297	masks: (list of) 2D arrays, int
298	masks output from Cellpose.eval, where 0=NO masks; 1,2,...=mask labels
299
300	flows: (list of) list of ND arrays
301	flows output from Cellpose.eval
302
303	file_names: (list of) str
304	names of files of images
305
306	"""
307
308	if isinstance(masks, list):	×
309	for image, mask, flow, file_name in zip(images, masks, flows, file_names):	×
310	save_masks(image, mask, flow, file_name, png=png, tif=tif, outlines=outlines)	×
311	return	×
312
313	if masks.ndim > 2 and not tif:	×
314	raise ValueError('cannot save 3D outputs as PNG, use tif option instead')	×
315	base = os.path.splitext(file_names)[0]	×
316	exts = []	×
317	if masks.ndim > 2 or masks.max()>2**16-1:	×
318	png = False	×
319	tif = True	×
320	if png:	×
321	exts.append('.png')	×
322	if tif:	×
323	exts.append('.tif')	×
324
325	# convert to uint16 if possible so can save as PNG if needed
326	masks = masks.astype(np.uint16) if masks.max()<2**16-1 else masks.astype(np.uint32)	×
327
328	# save masks
329	with warnings.catch_warnings():	×
330	warnings.simplefilter("ignore")	×
331	for ext in exts:	×
332	imsave(base + '_cp_masks' + ext, masks)	×
333
334	if png and MATPLOTLIB and not min(images.shape) > 3:	×
335	img = images.copy()	×
336	if img.ndim<3:	×
337	img = img[:,:,np.newaxis]	×
338	elif img.shape[0]<8:	×
339	np.transpose(img, (1,2,0))	×
340
341	fig = plt.figure(figsize=(12,3))	×
342	# can save images (set save_dir=None if not)
343	plot.show_segmentation(fig, img, masks, flows[0])	×
344	fig.savefig(base + '_cp_output.png', dpi=300)	×
345	plt.close(fig)	×
346
347	if masks.ndim < 3 and outlines:	×
348	mask_outlines = utils.outlines_list(masks)	×
349	outlines_to_text(base, mask_outlines)	×
350
351	def save_server(parent=None, filename=None):	3✔
352	""" Uploads a *_seg.npy file to the bucket.
353
354	Parameters
355	----------------
356
357	parent: PyQt.MainWindow (optional, default None)
358	GUI window to grab file info from
359
360	filename: str (optional, default None)
361	if no GUI, send this file to server
362
363	"""
364	if parent is not None:	×
365	q = QtGui.QMessageBox.question(	×
366	parent,
367	"Send to server",
368	"Are you sure? Only send complete and fully manually segmented data.\n (do not send partially automated segmentations)",
369	QtGui.QMessageBox.Yes \| QtGui.QMessageBox.No
370	)
371	if q != QtGui.QMessageBox.Yes:	×
372	return	×
373	else:
374	filename = parent.filename	×
375
376	if filename is not None:	×
377	os.environ['GOOGLE_APPLICATION_CREDENTIALS'] = os.path.join(os.path.dirname(os.path.realpath(__file__)),	×
378	'key/cellpose-data-writer.json')
379	bucket_name = 'cellpose_data'	×
380	base = os.path.splitext(filename)[0]	×
381	source_file_name = base + '_seg.npy'	×
382	io_logger.info(f'sending {source_file_name} to server')	×
383	time = datetime.datetime.now().strftime("%Y_%m_%d_%H_%M_%S.%f")	×
384	filestring = time + '.npy'	×
385	io_logger.info(f'name on server: {filestring}')	×
386	destination_blob_name = filestring	×
387	storage_client = storage.Client()	×
388	bucket = storage_client.bucket(bucket_name)	×
389	blob = bucket.blob(destination_blob_name)	×
390
391	blob.upload_from_filename(source_file_name)	×
392
393	io_logger.info(	×
394	"File {} uploaded to {}.".format(
395	source_file_name, destination_blob_name
396	)
397	)

MouseLand / cellpose / 149

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous