-
Notifications
You must be signed in to change notification settings - Fork 0
/
pendel.py
299 lines (271 loc) · 11 KB
/
pendel.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
# -*- coding: utf-8 -*-
#
# Copyright 2015: Manuel Schottdorf, Frank Stollmeier
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
__version__ = 0.5
import cv2
import numpy as np
from scipy import ndimage
import threading
from matplotlib import animation
import colorsys
import pylab as pl
import pdb
from time import time
pl.ion()
cap = cv2.VideoCapture(1)
def find_position(mask):
labeled_mask,num_label = ndimage.label(mask)
objects = ndimage.find_objects(labeled_mask)
object_sizes = [(slice1.stop-slice1.start)*(slice2.stop-slice2.start) for slice1,slice2 in objects]
biggest_object = objects[np.argmax(object_sizes)]
x,y = int(0.5*(biggest_object[0].start+biggest_object[0].stop)), int(0.5*(biggest_object[1].start+biggest_object[1].stop))
return x,y
positions_for_calibration = []
def set_center_pos(event,x,y,flags,param):
if event==cv2.EVENT_LBUTTONDBLCLK:
positions_for_calibration.append(np.array([x,y]))
print 'recieved coordinates: ' + repr([x,y])
center = np.array([0,0])
roi = [None,None,None,None]
color_tolerance = 5
saturation_tolerance = 100
value_tolerance = 100
t0 = time() #ref time
colors = []
color1range = []
color2range = []
shape = []
frame_rate = [None]
shift_color = lambda color,factor: np.array([np.clip(color[0]+factor*color_tolerance,0,255),np.clip(color[1]+factor*saturation_tolerance,0,255),np.clip(color[2]+factor*value_tolerance,0,255)],dtype=np.uint8)
circle_mask = None
zeroframe = None
def calibration():
'''Show the video from the webcam and set position of the center, color of the first marker and color of the second marker chosen by the user with three consecutive double clicks'''
cv2.namedWindow('calibration')
#reset previous calibartion values
if len(positions_for_calibration)==2:
for i in range(2):
shape.pop()
positions_for_calibration.pop()
colors.pop()
color1range.pop()
color2range.pop()
print 'Calibration: please doubleclick on the center, the first marker and the second marker'
cv2.setMouseCallback('calibration',set_center_pos)
time_start = time()
frames = 0
while True:
_, frame = cap.read()
frames += 1
cv2.imshow('calibration',frame)
k = cv2.waitKey(5) & 0xFF
if k == 27:
break
if len(positions_for_calibration)==3:
break
frame_rate[0] = frames / float(time() - time_start)
shape.append(frame.shape[0])
shape.append(frame.shape[1])
center[0] = positions_for_calibration[0][0]
center[1] = positions_for_calibration[0][1]
x,y = positions_for_calibration[1]
hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
color1 = hsv[y][x]
colors.append(color1)
color1range.append(shift_color(color1,-1))
color1range.append(shift_color(color1,+1))
x,y = positions_for_calibration[2]
hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
color2 = hsv[y][x]
colors.append(color2)
color2range.append(shift_color(color2,-1))
color2range.append(shift_color(color2,+1))
#select region of interest
v = np.array([x,y]) - center
r = np.sqrt(np.dot(v,v))
r = 1.1 * r
roi[0] = int(center[0]-r)
roi[1] = int(center[0]+r)
roi[2] = int(center[1]-r)
roi[3] = int(center[1]+r)
print "calibration complete"
if np.any(np.array(roi)<0):
print "warning: chosen ROI exceeds frame", repr(roi)
croi = np.clip(np.array(roi),0,np.inf)
roi[0],roi[1],roi[2],roi[3] = int(croi[0]),int(croi[1]),int(croi[2]),int(croi[3])
cv2.destroyWindow('calibration')
#circle mask
global circle_mask
xc,yc = np.meshgrid(np.arange(frame.shape[0]),np.arange(frame.shape[1]))
circle_mask = (xc-center[1])**2+(yc-center[0])**2 < r**2
circle_mask = np.transpose(np.array([circle_mask,circle_mask,circle_mask]))
global zeroframe
zeroframe = np.zeros(frame.shape,dtype=np.uint8)
return None
def capture(rotate_frames, show_roi):
'''capture one frame, locate the markers and return position'''
_, frame = cap.read()
hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
lower_color1,upper_color1 = color1range
lower_color2,upper_color2 = color2range
# Threshold the HSV image
mask_color1 = cv2.inRange(hsv, lower_color1, upper_color1)
mask_color2 = cv2.inRange(hsv, lower_color2, upper_color2)
#locate markers
x1,y1 = find_position(mask_color1)
cv2.circle(frame, (y1, x1), 2, (255, 255, 255), 20)
x2,y2 = find_position(mask_color2)
cv2.circle(frame, (y2, x2), 2, (255, 255, 255), 20)
#display the image
#pdb.set_trace()
frame = np.where(circle_mask,frame,zeroframe)
if rotate_frames:
phi1,phi2 = to_phase_coordinate(np.array([y1,x1]),np.array([y2,x2]))
R = cv2.getRotationMatrix2D(tuple(center),-phi1*180/np.pi,1)
frame = cv2.warpAffine(frame,R,(frame.shape[1],frame.shape[0]))
if show_roi:
frame = frame[roi[2]:roi[3],roi[0]:roi[1]]
cv2.imshow('tracker',frame)
k = cv2.waitKey(5) & 0xFF
if k == 27:
return x1,y1,x2,y2,True
else:
return x1,y1,x2,y2,False
def to_phase_coordinate(p1,p2):
'''convert positions p1 and p2 to angles phi1 and ph2'''
v1 = p1-center
v2 = p2-p1
phi1 = np.arctan2(v1[0],float(v1[1]))
phi2 = np.arctan2(v2[0],float(v2[1])) + phi1
return phi1,phi2
class Tracking(threading.Thread):
def __init__(self, rotate_frames = False, show_roi = True):
cv2.namedWindow('tracker')
self.rotate_frames = rotate_frames
self.show_roi = show_roi
threading.Thread.__init__(self)
def run(self):
self.time = 0
self.time_sec = np.zeros(1000000)
self.trajectory1 = np.zeros((2,1000000))
self.trajectory2 = np.zeros((2,1000000))
self.ptrajectory1 = np.zeros((2,1000000))
self.ptrajectory2 = np.zeros((2,1000000))
self.wtrajectory1 = np.zeros((2,1000000))
self.wtrajectory2 = np.zeros((2,1000000))
self.t_old = 0
while True:
try:
x1,y1,x2,y2,end = capture(self.rotate_frames, self.show_roi)
p1, p2 = to_phase_coordinate(np.array([x1, y1]),np.array([x2,y2]))
self.ptrajectory1[:,self.time] = p1
self.ptrajectory2[:,self.time] = p2
self.wtrajectory1[:,self.time] = (p1 - self.ptrajectory1[:,(self.time-1)]) / (time() - self.t_old)
self.wtrajectory2[:,self.time] = (p2 - self.ptrajectory2[:,(self.time-1)]) / (time() - self.t_old)
self.trajectory1[:,self.time] = y1,shape[0]-x1
self.trajectory2[:,self.time] = y2,shape[0]-x2
self.time_sec[self.time] = time() - t0
self.time += 1
self.t_old = time()
if end:
cv2.destroyWindow('tracker')
break
except ValueError:
print "Oh no, I've lost the marker!"
class Plot(object):
def __init__(self, track, frames=1000,interval=100):
self.fig = pl.figure()
ax = self.fig.add_subplot(231, autoscale_on=False, xlim=(0,shape[1]), ylim=(0,shape[0]))
ax.grid()
line0, = ax.plot([center[0]],[shape[0]-center[1]],'go',markersize=5)
line1, = ax.plot([np.nan],[np.nan],'-', color = colorsys.hsv_to_rgb(*(colors[0]/255.0)))
line1a, = ax.plot([np.nan],[np.nan],'o', color = colorsys.hsv_to_rgb(*(colors[1]/255.0)))
line1b, = ax.plot([np.nan],[np.nan],'-', color = colorsys.hsv_to_rgb(*(colors[0]/255.0)))
line1c, = ax.plot([np.nan],[np.nan],'o', color = colorsys.hsv_to_rgb(*(colors[1]/255.0)))
ax5 = self.fig.add_subplot(232, autoscale_on=False, xlim=(0, 5), ylim=(0, 500))
ax5.grid()
line5, = ax5.plot([], [], 'b-', lw=1)
line5a, = ax5.plot([], [], 'bo', lw=2)
line5b, = ax5.plot([], [], 'g-', lw=1)
line5c, = ax5.plot([], [], 'go', lw=2)
ax5.set_xlabel('Time [sec]')
ax5.set_ylabel('Position [px]')
ax6 = self.fig.add_subplot(233, autoscale_on=False, xlim=(0, 15), ylim=(0, 3000))
ax6.grid()
line6, = ax6.plot([], [], 'b.', lw=1)
ax6.set_xlabel('Frequency [Hz]')
ax6.set_ylabel('Amplitude')
ax2 = self.fig.add_subplot(234, autoscale_on=False, xlim=(0, 500), ylim=(-50, 50))
ax2.grid()
line2, = ax2.plot([], [], 'b.', lw=0.5)
line2a, = ax2.plot([], [], 'ro', lw=2)
ax2.set_xlabel('x-Pos 1')
ax2.set_ylabel('Angular Speed 1')
ax4 = self.fig.add_subplot(235, autoscale_on=False, xlim=(0,200), ylim=(0, 500))
ax4.grid()
line4, = ax4.plot([], [], 'ro', lw=0.2)
ax4.set_xlabel('Pos x1')
ax4.set_ylabel('Pos y1')
ax3 = self.fig.add_subplot(236, autoscale_on=False, xlim=(0,500), ylim=(-100,100))
ax3.grid()
line3, = ax3.plot([], [], 'b.', lw=0.5)
line3a, = ax3.plot([], [], 'ro', lw=2)
ax3.set_xlabel('x-Pos 2')
ax3.set_ylabel('Angular Speed 2')
phi1_old = 0 # for derivatives
phi2_old = 0
ls = [] # list for poincare section
def update(i):
x1 = track.trajectory1[0,:track.time]
y1 = track.trajectory1[1,:track.time]
x2 = track.trajectory2[0,:track.time]
y2 = track.trajectory2[1,:track.time]
phi1 = track.ptrajectory1[0,:track.time]
phi2 = track.ptrajectory2[0,:track.time]
w1 = track.wtrajectory1[0,:track.time]
w2 = track.wtrajectory2[0,:track.time]
time_elapsed = track.time_sec[0:track.time]
if i>10:
line1.set_data(x1,y1)
line1a.set_data(x1[-1],y1[-1])
line1b.set_data(x2,y2)
line1c.set_data(x2[-1],y2[-1])
line2.set_data(x1, w1)
line2a.set_data(x1[-1], w1[-1])
line3.set_data(x2, w2)
line3a.set_data(x2[-1], w1[-1])
if ((w1[-1] > 0) and (w1[-2] < 0) ) or ((w1[-1] < 0) and (w1[-2] > 0) ):
ls.append([x1[-1],y1[-1]])
a = np.array(ls)
line4.set_data(a[:,0],a[:,1])
if i>100:
tt = time_elapsed[(-101):(-1)] - time_elapsed[-101]
line5.set_data(tt, y1[(-101):(-1)])
line5a.set_data(time_elapsed[-1], y1[-1])
line5b.set_data(tt,y2[(-101):(-1)])
line5c.set_data(time_elapsed[-1], y2[-1])
spec = np.fft.fft(y1[(-101):(-1)])
timestep = np.mean(np.diff(tt))
freq = np.fft.fftfreq(spec.size, d=timestep)
sp = np.abs(spec)
line6.set_data(freq,sp)
return [line1,line1a, line1b, line1c, line2, line2a, line3, line3a, line5, line5a, line5b, line5c, line6]
self.animated_pendulum = animation.FuncAnimation(self.fig, update, frames=frames,interval=interval, blit = True)
calibration()
#x1,y1,x2,y2,end,frame = capture(True, False)
track = Tracking(rotate_frames=True, show_roi = False)
track.start()
#plot = Plot(track)