-
Notifications
You must be signed in to change notification settings - Fork 0
/
skimageApi.py
524 lines (384 loc) · 13.6 KB
/
skimageApi.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
# -*- encoding: utf-8 -*-
'''
@Author : lance
@Email : [email protected]
'''
from skimage import io,transform,exposure
import numpy as np
import os
imgpath="./data/test/C3F/C3F_blockId#32756.bmp"
img=io.imread(imgpath)
io.imshow(img)
print(type(img)) #显示类型
print(img.shape) #显示尺寸 #height/weight/channel
print(img.size) #显示总像素个数
print(img.max()) #最大像素值
print(img.min()) #最小像素值
print(img.mean()) #像素平均值
#保存图片
#io.imsave('d:/cat.jpg',img)
#随机生成5000个椒盐
rows,cols,dims=img.shape
for i in range(500000):
x=np.random.randint(0,rows)
y=np.random.randint(0,cols)
img[x,y,:]=255
io.imshow(img)
#二值化
#颜色空间转换 skimage.color.convert_colorspace(arr, fromspace, tospace)
from skimage import color
img_gray=color.rgb2gray(img)
rows,cols=img_gray.shape
for i in range(rows):
for j in range(cols):
if (img_gray[i,j]<=0.5):
img_gray[i,j]=0
else:
img_gray[i,j]=1
io.imshow(img_gray)
print(img_gray[img_gray>0].size) #白色区域的像素个数 - 面积
index=np.where(img_gray==1)
h1=index[0].min()
h2=index[0].max()
w1=index[1].min()
w2=index[1].max()
roi=img[h1:h2,w1:w2]
io.imshow(roi)
#对于r通道大于170的像素值进行过滤后显示G通道为255
reddish = img[:, :, 0] >170
img[reddish] = [0, 255, 0]
io.imshow(img)
#分类着色
gray=color.rgb2gray(img)
rows,cols=gray.shape
labels=np.zeros([rows,cols])
for i in range(rows):
for j in range(cols):
if(gray[i,j]<0.4):
labels[i,j]=0
elif(gray[i,j]<0.75):
labels[i,j]=1
else:
labels[i,j]=2
dst=color.label2rgb(labels)
io.imshow(dst)
#批处理
str= './data/test/C3F/*.bmp'
def convert_gray(f):
rgb=io.imread(f) #依次读取rgb图片
gray=color.rgb2gray(rgb) #将rgb图片转换成灰度图
dst=transform.resize(gray,(256,256)) #将灰度图片大小转换为256*256
return dst
coll = io.ImageCollection(str,load_func=convert_gray)
path= "./data/测试/"
if not os.path.exists(path):
os.makedirs(path)
for i in range(len(coll)):
io.imsave(path+np.str(i)+'.jpg',coll[i]) #循环保存图片
#图像滤波
from skimage import filters
import matplotlib.pyplot as plt
#转换为灰度空间
img=color.rgb2gray(img)
#edges = filters.sobel(img) #边缘检测
#edges = filters.roberts(img) #边缘检测
edges = filters.gaussian(img,sigma=5) #消除噪音 sigma越小越清晰
plt.imshow(edges,cmap="gray")
plt.show()
#图像自动阈值分割
from skimage import filters
import matplotlib.pyplot as plt
img=color.rgb2gray(img)
#thresh = filters.threshold_otsu(img) #返回一个阈值
thresh = filters.threshold_li(img)
dst =(img <= thresh)*1.0 #根据阈值进行分割
plt.figure('thresh',figsize=(8,8))
plt.subplot(121)
plt.title('original image')
plt.imshow(img,plt.cm.gray)
plt.subplot(122)
plt.title('binary image')
plt.imshow(dst,plt.cm.gray)
plt.show()
#阈值分割
from skimage import filters
import matplotlib.pyplot as plt
image =color.rgb2gray(img)
dst =filters.threshold_local(image, 15) #返回一个阈值图像
plt.figure('thresh',figsize=(8,8))
plt.subplot(121)
plt.title('original image')
plt.imshow(image,plt.cm.gray)
plt.subplot(122)
plt.title('binary image')
plt.imshow(dst,plt.cm.gray)
plt.show()
#膨胀 用来扩充边缘或填充小的孔洞。
import skimage.morphology as sm
import matplotlib.pyplot as plt
img=color.rgb2gray(img)
dst1=sm.dilation(img,sm.square(5)) #用边长为5的正方形滤波器进行膨胀滤波
dst2=sm.dilation(img,sm.square(15)) #用边长为15的正方形滤波器进行膨胀滤波
plt.figure('morphology',figsize=(8,8))
plt.subplot(131)
plt.title('origin image')
plt.imshow(img,plt.cm.gray)
plt.subplot(132)
plt.title('morphological image')
plt.imshow(dst1,plt.cm.gray)
plt.subplot(133)
plt.title('morphological image')
plt.imshow(dst2,plt.cm.gray)
#腐蚀 可用来提取骨干信息，去掉毛刺，去掉孤立的像素。
import skimage.morphology as sm
import matplotlib.pyplot as plt
img=img=color.rgb2gray(img)
dst1=sm.erosion(img,sm.square(5)) #用边长为5的正方形滤波器进行膨胀滤波
dst2=sm.erosion(img,sm.square(25)) #用边长为25的正方形滤波器进行膨胀滤波
plt.figure('morphology',figsize=(8,8))
plt.subplot(131)
plt.title('origin image')
plt.imshow(img,plt.cm.gray)
plt.subplot(132)
plt.title('morphological image')
plt.imshow(dst1,plt.cm.gray)
plt.subplot(133)
plt.title('morphological image')
plt.imshow(dst2,plt.cm.gray)
#开运算 先腐蚀再膨胀 消除小物体
from skimage import io,color
import skimage.morphology as sm
import matplotlib.pyplot as plt
img=color.rgb2gray(img)
dst=sm.opening(img,sm.disk(9)) #用边长为9的圆形滤波器进行膨胀滤波
plt.figure('morphology',figsize=(8,8))
plt.subplot(121)
plt.title('origin image')
plt.imshow(img)
plt.axis('off')
plt.subplot(122)
plt.title('morphological image')
plt.imshow(dst,plt.cm.gray)
plt.axis('off')
#闭运算 先膨胀再腐蚀 填充空洞
from skimage import io,color
import skimage.morphology as sm
import matplotlib.pyplot as plt
img=color.rgb2gray(img)
dst=sm.closing(img,sm.disk(9)) #用边长为5的圆形滤波器进行膨胀滤波
plt.figure('morphology',figsize=(8,8))
plt.subplot(121)
plt.title('origin image')
plt.imshow(img,plt.cm.gray)
plt.axis('off')
plt.subplot(122)
plt.title('morphological image')
plt.imshow(dst,plt.cm.gray)
plt.axis('off')
#%%
import matplotlib.pyplot as plt
from skimage.segmentation import mark_boundaries
from skimage.morphology import extrema
from scipy import linalg
import numpy as np
from skimage import io
from skimage import img_as_float
from skimage.color import rgb2gray
from skimage import exposure
from skimage import color
from scipy import ndimage as ndi
from skimage.morphology import label, binary_erosion, binary_dilation, square, binary_opening, binary_closing
import operator
from itertools import repeat
fn='E:\Python\ANN\Projects\固高\定位\Image1.jpg'
img= io.imread(fn)
# using functions in Skimage #
# note the input image is assumed to be sRGB, and will be normalized to [0,1] range #
img_xyz= color.rgb2xyz(img)
img_lab= color.xyz2lab(img_xyz)
#plt.imshow(img_lab, cmap="gray")
#plt.show()
# using VN boy's coversion method #
# assume the input is CIE RGB #
def rgb2xyz_vn(img):
img= img.copy()/255.
conv_mat = np.array([[ 0.4900, 0.3100, 0.2000],
[ 0.1769, 0.8124, 0.0107],
[ 0.0000, 0.0099, 0.9901]])
xyz = np.dot(img, conv_mat.T)
return xyz
img_xyz_vn= rgb2xyz_vn(img)
# convert the xyz to lab:
# according to http://www.brucelindbloom.com/index.html?Eqn_RGB_XYZ_Matrix.html
# the reference white for CIE RGB is E
img_lab_vn= color.xyz2lab(img_xyz_vn, illuminant= "E")
gray= img_lab_vn[:, :, 0]
#gray = rgb2gray(img)*255
#gray = gray.astype(int)
gray = rgb2gray(img)
#gray = exposure.equalize_hist(gray)
plt.imshow(gray, cmap="gray")
plt.show()
############################################################################
########### Ostu thresholding #############
import numpy as np
from skimage import img_as_float, filters, exposure
from skimage.filters import gaussian, laplace
from skimage.exposure import rescale_intensity
from skimage import io
import cv2
img_h= gray
blur= gaussian(img_h, sigma= 1)
laplacian = laplace(blur)
theCutoff= np.percentile(abs(laplacian).flatten(), 80)
laplacian_index= abs(laplacian)> theCutoff
strong_pixels= blur[laplacian_index]
blur[laplacian_index]= 255
blur[~laplacian_index]= 0
laplacian_list= laplacian.flatten()
## Use skimage to perform otsu thresholding ##
val = filters.threshold_otsu(strong_pixels)
plt.hist(strong_pixels[strong_pixels<val], bins=50)
plt.show()
hist, bins_center = exposure.histogram(strong_pixels)
plt.imshow(img_h>val, cmap='gray', interpolation='nearest')
#plt.axis('off')
plt.show()
###########
image= img_h>val
image= ndi.morphology.binary_fill_holes(image)
image= binary_opening(image, selem=square(3))
image= binary_closing(image, selem=square(3))
image= binary_erosion(image, selem= square(5))
image= binary_dilation(image, selem= square(5))
plt.imshow(image, cmap='gray', interpolation='nearest')
#plt.axis('off')
plt.show()
theConnected, theConnected_num= label(image, return_num=True)
theConnected_ind= np.unique(theConnected).tolist()
del theConnected_ind[theConnected_ind.index(0)]
theConnected_num_P= []
for i in np.unique(theConnected_ind):
the_Ps= image[theConnected== i]
theConnected_num_P.append(len(the_Ps))
# Order the connected components by the number of pixels #
new_theConnected_num_P= np.array(theConnected_num_P).copy().tolist()
new_theConnected_num_P.sort(reverse=True)
cand_max= np.where(np.array(theConnected_num_P)== new_theConnected_num_P[0])[0]
cand_2= np.where(np.array(theConnected_num_P)== new_theConnected_num_P[1])[0]
cand_3= np.where(np.array(theConnected_num_P)== new_theConnected_num_P[2])[0]
newImg= np.zeros(image.shape, dtype= "uint8")
newImg[theConnected==theConnected_ind[cand_max[0]]]=1
newImg[theConnected==theConnected_ind[cand_2[0]]]=2
newImg[theConnected==theConnected_ind[cand_3[0]]]=3
plt.imshow(newImg, cmap='gray', interpolation='nearest')
plt.show()
fig01= plt.imshow(newImg, cmap='gray', interpolation='nearest')
plt.xticks([]), plt.yticks([])
plt.tight_layout()
fig01.figure.savefig(samLoc + samName + 'thresholded.png', dpi=200)
############################
### Boundary tracing ###
# obtain the coordinates of each center #
cen_max= (int(np.median(np.where(newImg==1)[0])), int(np.median(np.where(newImg==1)[1])))
cen_2= (int(np.median(np.where(newImg==2)[0])), int(np.median(np.where(newImg==2)[1])))
cen_3= (int(np.median(np.where(newImg==3)[0])), int(np.median(np.where(newImg==3)[1])))
cen_max2= (cen_max[0], cen_2[1])
cen_max3= (cen_max[0], cen_3[1])
all_bounds= []
for sz in [1,2,3]:
oneLumen= newImg==sz
eroded= binary_erosion(oneLumen, selem= square(3))
boundary_01= oneLumen ^ eroded
boundary= np.zeros(tuple(map(operator.add, boundary_01.shape, (2,2))), dtype= "uint8")
boundary[1:(boundary.shape[0]-1), 1:(boundary.shape[1]-1)]= boundary_01
boundary= boundary>0
plt.imshow(boundary)
plt.show()
### Sample lumen boundary pixels with roughly equal intervals ###
# 1) trace the boundary in clockwise direction #
theX, theY= np.where(boundary)
totalNum= range(len(theX))
n_s= [(0,-1), (-1,-1), (-1,0), (-1,1), (0,1), (1,1), (1,0), (1,-1)]
k= 0
bound_pixels= [(theX[0], theY[0])]
for i in totalNum:
if i>0:
if bound_pixels[i]== bound_pixels[0]:
break
if k != 0:
ind= range(k, 8)+ range(k)
tmp_n_s= [n_s[x] for x in ind]
else:
tmp_n_s= n_s
for j in range(8):
back_xy= tuple(np.add(bound_pixels[i], tmp_n_s[j]))
if boundary[back_xy]:
bound_pixels.append(back_xy)
back_xy_last= tuple(np.add(bound_pixels[i], tmp_n_s[j-1]))
#k= n_s.index(tmp_n_s[j-1])
k= n_s.index(tuple(np.subtract(back_xy_last, back_xy)))
break
bound_pixels= bound_pixels[:-1]
bound_pixels= [(a[0]-1, a[1]-1) for a in bound_pixels]
# 2) sample the boundary pixels at equal intervals #
interval= 1
num_elements= len(bound_pixels)/ interval
if len(bound_pixels) % interval < 5:
num_elements= num_elements
else:
num_elements= num_elements+1
bp_index= [interval* x for x in range(num_elements)]
lbp= [bound_pixels[x] for x in bp_index]
all_bounds.append(lbp)
# visually test the result #
for szh in range(len(all_bounds)):
lbp= all_bounds[szh]
frame= np.zeros(boundary.shape, dtype= "uint8")
for i in range(len(lbp)):
frame[lbp[i][0], lbp[i][1]]=1
fig01= plt.imshow(frame, cmap='gray')
plt.show()
# Find the closest boundary pixels to the bar center #
theData= all_bounds[1]
theCenter= cen_max2
theDis= []
for k in range(len(theData)):
lu_x= theData[k][0]
lu_y= theData[k][1]
dis_01= np.round(np.sqrt((lu_x- theCenter[0])**2 + (lu_y- theCenter[1])**2), 2)
theDis.append(dis_01)
theCand= np.where(np.array(theDis)== np.min(theDis))[0]
tip_1= theData[theCand[0]]
theData= all_bounds[2]
theCenter= cen_max3
theDis= []
for k in range(len(theData)):
lu_x= theData[k][0]
lu_y= theData[k][1]
dis_01= np.round(np.sqrt((lu_x- theCenter[0])**2 + (lu_y- theCenter[1])**2), 2)
theDis.append(dis_01)
theCand= np.where(np.array(theDis)== np.min(theDis))[0]
tip_2= theData[theCand[0]]
# Place the tip point coordinates into the original figure #
tip_1_x= tip_1[0]+ range(-30, 30)
tip_1_y= tip_1[1]+ range(-30, 30)
tip_2_x= tip_2[0]+ range(-30, 30)
tip_2_y= tip_2[1]+ range(-30, 30)
#frame= np.zeros(gray.shape, dtype= "uint8")
#frame[tip_1_x, np.array([tip_1[1]]*len(tip_1_x))]=1
#frame[np.array([tip_1[0]]*len(tip_1_y)), tip_1_y]=1
#frame[tip_2_x, np.array([tip_2[1]]*len(tip_2_x))]= 1
#frame[[tip_2[0]]*len(tip_2_y), tip_2_y.tolist()]= 1
#fig01= plt.imshow(frame, cmap='gray', interpolation='nearest')
#plt.xticks([]), plt.yticks([])
#plt.tight_layout()
#fig01.figure.savefig(samLoc + samName + 'barTipAdded.png', dpi=200)
img[tip_1_x, np.array([tip_1[1]]*len(tip_1_x))]= [0, 0, 255]
img[np.array([tip_1[0]]*len(tip_1_y)), tip_1_y]= [0, 0, 255]
img[tip_2_x, np.array([tip_2[1]]*len(tip_2_x))]= [255, 0, 0]
img[np.array([tip_2[0]]*len(tip_2_y)), tip_2_y]= [255, 0, 0]
fig01= plt.imshow(img)
#plt.xticks([]), plt.yticks([])
plt.tight_layout()
fig01.figure.savefig(samLoc + samName + 'barTipAdded.png', dpi=200)