-
Notifications
You must be signed in to change notification settings - Fork 0
/
objsplit_argp_v2.py
260 lines (225 loc) · 11.6 KB
/
objsplit_argp_v2.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
# coding: utf-8
# OBJECT SPLITTER (OBJSPLIT) - ARGPARSED VERSION -
# -----------------------------------------------------------------------------------------------------------------------------
# Author: Antonio Oliver Gelabert (ORCID : http://orcid.org/0000-0001-8571-2733)
# -----------------------------------------------------------------------------------------------------------------------------
# Parameters
# -----------------------------------------------------------------------------------------------------------------------------
# sc : scale conversion (in units/pixel)
# tmin : minimum binary threshold for object detection
# tmax : maximum binary threshold for object detection
# fmin : minimum area filter (in units**2)
# fmax : maximum area filter (in units**2)
# f : filename of the input file
# split : Split or not in separated outputs the objects found
# thm : Binarization method (B: simple binary, G: gaussian, O: Otsu). Default: Gaussian (G))
# inv : Invert binary mask (y/n)
# -----------------------------------------------------------------------------------------------------------------------------
from datetime import datetime
import time as tm
import time
import os
import sys
import cv2
import numpy as np
import matplotlib.pyplot as plt
import warnings
import argparse
t0= tm.clock()
# Parsing optional arguments of the program
ap = argparse.ArgumentParser()
ap.add_argument("-sc", "--scale", required=False, default='1', help="scale dymensions. Default value: 1 unit/pixel")
ap.add_argument("-tmin", "--tmin", required=False, default='150', help="Low threshold brightness (between 0 and 255, close to 0 and bigger, to exclude dark images). Default value:150")
ap.add_argument("-tmax", "--tmax", required=False, default='255', help="High threshold brightness (between 0 and 255, close to 0 and bigger, to exclude dark images). Default value:255")
ap.add_argument("-fmin", "--filtmin", required=False, default='10', help="Mininum area filter (in units**2). Default value:10 pixels")
ap.add_argument("-fmax", "--filtmax", required=False, default='99999999', help="maximum area filter (in units**2). Default value:no limit")
ap.add_argument("-f", "--filename", required=True, help="Filename of input images. ")
ap.add_argument("-split", "--split", required=False, default='n', help="Split (y) or not (n) in separated outputs the objects found. Default: no")
ap.add_argument("-thm", "--thresholdmethod", required=False, default='G', help="Binarization method (B: simple binary, G: gaussian, O: Otsu). Default: Gaussian (G))")
ap.add_argument("-inv", "--invert", required=False, default='n', help="Invert binary mask (y/n). Default: No")
args = vars(ap.parse_args())
# Assign args to program variables
fname_sb=args["filename"]
scale=int(args["scale"])
tresh_8b_min=int(args["tmin"])
tresh_8b_max=int(args["tmax"])
filtareamin=int(args["filtmin"])/(scale*scale)
filtareamax=int(args["filtmax"])/(scale*scale)
splitb=args["split"]
thrmethod=args["thresholdmethod"]
invert=args["invert"]
dirstr=str(round(time.time()))
os.mkdir("out_"+dirstr)
if(splitb == "y"):
os.mkdir("out_"+dirstr+"/split")
f2 = open("out_"+dirstr+"/"+dirstr+"_ind_data.txt", 'w')
f2.write('ID,L,W,Area,perimeter,xc,yc\n')
img = cv2.imread(fname_sb)
font = cv2.FONT_HERSHEY_SIMPLEX
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
if(thrmethod == 'B'):
if(invert == 'y'):
ret,patt = cv2.threshold(gray,tresh_8b_min,tresh_8b_max,cv2.THRESH_BINARY_INV)
else:
ret,patt = cv2.threshold(gray,tresh_8b_min,tresh_8b_max,cv2.THRESH_BINARY)
elif(thrmethod == 'G'):
if(invert =='y'):
patt = cv2.adaptiveThreshold(gray,tresh_8b_max,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY_INV,11,2)
else:
patt = cv2.adaptiveThreshold(gray,tresh_8b_max,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,11,2)
cv2.imwrite("out_"+dirstr+"/"+dirstr+"_binary.jpg", patt)
elif(thrmethod == 'O'):
blur = cv2.GaussianBlur(gray,(5,5),0)
if(invert =='y'):
ret,patt = cv2.threshold(blur,0,255,cv2.THRESH_BINARY_INV+cv2.THRESH_OTSU)
else:
ret,patt = cv2.threshold(blur,0,255,cv2.THRESH_BINARY+cv2.THRESH_OTSU)
cv2.imwrite("out_"+dirstr+"/"+dirstr+"_binary.jpg", patt)
contours, hierarchy = cv2.findContours(patt, cv2.RETR_LIST, cv2.CHAIN_APPROX_NONE)
contpic=cv2.imread(fname_sb)# cv2.drawContours(bwcont, contours, -1, (255, 0, 0), 1)
areatot=0
areafilt=0
nfilt=0
areafstats=[]
perfstats=[]
areastats=[]
boxc=[]
Lboxst=[]
Wboxst=[]
for j in range(0,len(contours),1):
area = cv2.contourArea(contours[j])*scale**2
perimeter = cv2.arcLength(contours[j],True)*scale
areatot=areatot+area
areastats.append(area)
if (area > filtareamin) & (area < filtareamax):
nfilt=nfilt+1
areafstats.append(area)
perfstats.append(perimeter)
areafilt=areafilt+area
rect = cv2.minAreaRect(contours[j])
box = cv2.boxPoints(rect)
box = np.int0(box)
boxc.append(box)
x0=box[0][0]
y0=box[0][1]
x1=box[1][0]
y1=box[1][1]
x2=box[2][0]
y2=box[2][1]
x3=box[3][0]
y3=box[3][1]
pm0x=int(round((box[0][0]+box[1][0])/2))
pm0y=int(round((box[0][1]+box[1][1])/2))
pm1x=int(round((box[1][0]+box[2][0])/2))
pm1y=int(round((box[1][1]+box[2][1])/2))
pm2x=int(round((box[2][0]+box[3][0])/2))
pm2y=int(round((box[2][1]+box[3][1])/2))
pm3x=int(round((box[3][0]+box[0][0])/2))
pm3y=int(round((box[3][1]+box[0][1])/2))
d01box=((box[0][0]-box[1][0])**2+(box[0][1]-box[1][1])**2)**0.5
d12box=((box[1][0]-box[2][0])**2+(box[1][1]-box[2][1])**2)**0.5
Abox=d01box*d12box*scale**2
Lbox=np.maximum(d01box,d12box)*scale
Wbox=np.minimum(d01box,d12box)*scale
Lboxst.append(Lbox)
Wboxst.append(Wbox)
xmaxb=np.max((x0,x1,x2,x3))
ymaxb=np.max((y0,y1,y2,y3))
xminb=np.min((x0,x1,x2,x3))
yminb=np.min((y0,y1,y2,y3))
if(xminb < 1):
xminb=1
if(yminb < 1):
yminb=1
if(xmaxb < 1):
xmaxb=1
if(ymaxb < 1):
ymaxb=1
#cv2.putText(contpic, str(round(area,2)), (int(round((pm0x+pm2x)*0.5)+5),int(round((pm2y+pm0y)*0.5))), font, 0.3, (255, 0, 0), 1, cv2.LINE_AA)
cv2.putText(contpic, str(nfilt), (int(round((pm0x+pm2x)*0.5)+5),int(round((pm2y+pm0y)*0.5))), font, 0.3, (255, 0, 0), 1, cv2.LINE_AA)
cv2.drawContours(contpic, contours[j], -1, (255, 0, 0), 1)
cv2.drawContours(contpic,[box],0,(0,255,255),1)
if(splitb == "y"):
cv2.imwrite("out_"+dirstr+"/split/filtobj_"+str(nfilt)+".jpg", contpic[yminb:ymaxb,xminb:xmaxb,:])
f2.write(str(nfilt)+','+str(round(Lbox,2))+','+str(round(Wbox,2))+','+str(round(area,2))+','+str(round((pm0x+pm2x)*0.5,2))+','+str(round((pm2y+pm0y)*0.5,2))+'\n')
cv2.imwrite("out_"+dirstr+"/"+dirstr+"_allcont.jpg", contpic)
f2.close()
# Initial set of parameters and statistical information
f3 = open("out_"+dirstr+"/"+dirstr+"_parameters_and_outputs.txt", 'w')
# Cabecera del programa
now = datetime.now()
year = now.strftime("%Y")
month = now.strftime("%m")
day = now.strftime("%d")
time = now.strftime("%H:%M:%S")
f3.write("************************************************************************************************************"+'\n')
f3.write("************************************************************************************************************"+'\n')
f3.write("This ouput has been generated on " + day +"th of " + month + " of year " + year + ", at local time: " + time +'\n')
f3.write("************************************************************************************************************"+'\n')
f3.write("************************************************************************************************************"+'\n')
f3.write("OBJECT DETECTOR&COUNTER IN IMAGES"+'\n')
f3.write("************************************************************************************************************"+'\n')
f3.write("************************************************************************************************************"+'\n')
f3.write("Author : Antonio Oliver Gelabert"+'\n')
f3.write("Contact : [email protected]"+'\n')
f3.write("************************************************************************************************************"+'\n')
f3.write("***************************** PARAMETERS **************************************************"+'\n')
f3.write("************************************************************************************************************"+'\n')
f3.write("Filename : "+fname_sb+'\n')
f3.write("Scale : "+str(scale)+" (reference in units/pixel)\n")
f3.write("Minimum intensity threshold : "+str(tresh_8b_min)+'\n')
f3.write("Maximum intensity threshold : "+str(tresh_8b_max)+'\n')
f3.write("Minimum area filter : "+str(filtareamin)+'\n')
f3.write("Maximum area filter :" +str(filtareamax)+'\n')
f3.write("************************************************************************************************************"+'\n')
f3.write("***************************** OUTPUTS **************************************************"+'\n')
f3.write("************************************************************************************************************"+'\n')
f3.write("Number of objects found: "+str(len(contours))+'\n')
f3.write("Average leght of objects found (in units): "+ str(round(np.mean(Lboxst),2))+"+/-"+str(round(np.std(Lboxst),2))+'\n')
f3.write("Average width of objects found (in units): "+ str(round(np.mean(Wboxst),2))+"+/-"+str(round(np.std(Wboxst),2))+'\n')
f3.write("Average area of objects found (in units**2): "+ str(round(np.mean(areafstats),2))+"+/-"+str(round(np.std(areafstats),2))+'\n')
f3.write("Average perimeter of objects found (in units): "+ str(round(np.mean(perfstats),2))+"+/-"+str(round(np.std(perfstats),2))+'\n')
f3.write("************************************************************************************************************"+'\n')
f3.write("************************************************************************************************************"+'\n')
# Statistical analysis plots
# create a figure
fig = plt.figure()
# define subplots
plt1 = fig.add_subplot(221)
plt2 = fig.add_subplot(222)
plt3 = fig.add_subplot(223)
plt4 = fig.add_subplot(224)
plt1.hist(areafstats, bins=10)
plt1.set_title('Area')
plt2.hist(Lboxst, bins=10)
plt2.set_title('Lenght')
plt3.hist(Wboxst, bins=10)
plt3.set_title('Width')
plt4.hist(perfstats, bins=10)
plt4.set_title('Perimeter')
# Space between subplots
fig.subplots_adjust(hspace=.5,wspace=0.5)
plt.savefig("out_"+dirstr+"/"+dirstr+"_histplot.jpg",dpi=200)
# create a figure
fig2 = plt.figure()
# define subplots
plt5 = fig2.add_subplot(221)
plt6 = fig2.add_subplot(222)
plt7 = fig2.add_subplot(223)
plt8 = fig2.add_subplot(224)
plt5.boxplot(areafstats)
plt5.set_title('Area')
plt6.boxplot(Lboxst)
plt6.set_title('Lenght')
plt7.boxplot(Wboxst)
plt7.set_title('Width')
plt8.boxplot(perfstats)
plt8.set_title('Perimeter')
# Space between subplots
fig2.subplots_adjust(hspace=.5,wspace=0.5)
plt.savefig("out_"+dirstr+"/"+dirstr+"_boxsplot.jpg",dpi=200)
t1 = tm.clock()
f3.write("Job time (s) : "+ str(np.round(t1 - t0,2))+'\n')
f3.write("************************************************************************************************************"+'\n')
f3.write("************************************************************************************************************"+'\n')
f2.close()