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
|
import cv2
import math
import numpy
from math import sin,cos
cap = cv2.VideoCapture("../vid.mp4")
feature_params = dict( maxCorners = 100,
qualityLevel = 0.3,
minDistance = 20,
blockSize = 7 )
scale_factor=0.5
scr_width=1600
scr_height=900
# Parameters for lucas kanade optical flow
lk_params = dict( winSize = (15,15),
maxLevel = 2,
criteria = (cv2.TERM_CRITERIA_EPS | cv2.TERM_CRITERIA_COUNT, 10, 0.03))
ret,oldframe=cap.read()
oldgray=cv2.cvtColor(oldframe,cv2.COLOR_BGR2GRAY)
height, width, bpp = oldframe.shape
mask = numpy.ones((height,width, 1), numpy.uint8) * 255
screencontent = numpy.zeros((scr_height, scr_width,3), numpy.uint8)
total_angle=0.
total_x=1500
total_y=0
while(cap.isOpened()):
ret, frame = cap.read()
gray = cv2.cvtColor(frame,cv2.COLOR_BGR2GRAY)
p0 = cv2.goodFeaturesToTrack(oldgray, mask = None, **feature_params)
# calculate optical flow
p1, st, err = cv2.calcOpticalFlowPyrLK(oldgray, gray, p0, None, **lk_params)
# Select good points
good_new = p1[st==1]
good_old = p0[st==1]
for i in p0:
x,y=i.ravel()
cv2.line(frame, (x,y),(x,y), (0,0,255),5)
one=[]
two=[]
for i,(new,old) in enumerate(zip(good_new, good_old)):
x1,y1=new.ravel()
x2,y2=old.ravel()
cv2.line(frame, (x1,y1),(x2,y2), (0,255,0),5)
one=one+[(x1,y1)]
two=two+[(x2,y2)]
mat = cv2.estimateRigidTransform(gray,oldgray,False)
angle = math.atan2(mat[0,1],mat[0,0])
stretch = int((math.sqrt(mat[0,1]**2+mat[0,0]**2)-1)*100)
# calculate shift_x and _y is if one would rotate-and-stretch around the center of the image, not the topleft corner
shift_x = mat[0,2] - width/2 + ( mat[0,0]*width/2 + mat[0,1]*height/2 )
shift_y = mat[1,2] - height/2 + ( mat[1,0]*width/2 + mat[1,1]*height/2 )
total_x = total_x + shift_x
total_y = total_y + shift_y
total_angle=total_angle+angle
print angle/3.141592654*180,'deg\t',stretch,"%\t", shift_x,'\t',shift_y
mat2=cv2.getRotationMatrix2D((width/2,height/2), total_angle/3.141593654*180, scale_factor)
mat2[0,2] = mat2[0,2]+total_x*scale_factor
mat2[1,2] = mat2[1,2]+total_y*scale_factor
print mat2.__repr__()
# mat2=numpy.array([[cos(total_angle), sin(total_angle), total_x], [-sin(total_angle),cos(total_angle),total_y]])
frame2= cv2.warpAffine(frame, mat2, (scr_width,scr_height) )
mask2 = cv2.warpAffine(mask, mat2, (scr_width,scr_height) )
ret, mask2 = cv2.threshold(mask2, 254, 255, cv2.THRESH_BINARY)
mask2 = cv2.erode(mask2, numpy.ones((2,2),numpy.uint8)) # strip off the potentially-badlooking edges
#screencontent = frame2
screencontent = cv2.bitwise_and(screencontent,screencontent, mask=cv2.bitwise_not(mask2))
screencontent = cv2.add(screencontent, cv2.bitwise_and(frame2,frame2,mask=mask2))
#screencontent = frame2
#screencontent = mask2
cv2.imshow('frame', frame)
cv2.imshow('screencontent', screencontent)
oldframe=frame
oldgray=gray
if cv2.waitKey(20) & 0xFF == ord("q"):
break
|