structure-based-visual-localization/feat_matcher.py at master · nadiawangberg/structure-based-visual-localization · GitHub

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
import cv2 as cv
import numpy as np
from matplotlib import pyplot as plt

img1 = cv.imread('data/im1.png')
img2 = cv.imread('data/im2.png')
gray1= cv.cvtColor(img1,cv.COLOR_BGR2GRAY)
gray2= cv.cvtColor(img2,cv.COLOR_BGR2GRAY)

#Detectors
sift = cv.xfeatures2d.SIFT_create(0,3,0.04, 10,1.4)
orb = cv.ORB_create()

#should change params

#Detection - SIFT
kp1, des1 = sift.detectAndCompute(gray1,None) #Descriptor also computed
kp2, des2 = sift.detectAndCompute(gray2,None) #Descriptor also computed
#print(kp1[0].pt[0])

#orb for img1
kp1_o = orb.detect(gray1,None)
kp1_o, des1_o = orb.compute(gray1, kp1_o)

#orb for img2
kp2_o = orb.detect(gray2,None)
kp2_o, des2_o = orb.compute(gray2, kp2_o)

#FLANN - SIFT
FLANN_INDEX_KDTREE = 1
index_params = dict(algorithm = FLANN_INDEX_KDTREE, trees = 5)
search_params = dict(checks=50)   # or pass empty dictionary
flann = cv.FlannBasedMatcher(index_params,search_params)

#SIFT
matches = flann.knnMatch(des1,des2,k=2)
matches_filterd = []

#Try to only use good matches
matchesMask = [[0,0] for i in range(len(matches))]

#Ratio test (SIFT)
for i,(m,n) in enumerate(matches):
    if m.distance < 0.35*n.distance:
        matchesMask[i]=[1,0]
        matches_filterd.append([kp1[i].pt[0], kp1[i].pt[1], kp2[i].pt[0], kp2[i].pt[1]])


with open('matches_sift.txt', 'a') as f:
    for match in matches_filterd:
        for coordinate in match:
            f.write("%s " % coordinate) #write each number
        f.write("\n") #end with newline


draw_params = dict(
                   singlePointColor = (255,0,0),
                   matchesMask = matchesMask,
                   flags = cv.DrawMatchesFlags_DEFAULT)

img_sift = cv.drawMatchesKnn(gray1,kp1,gray2,kp2,matches,None,**draw_params)

plt.figure(1)
#plt.subplot(211)
plt.imshow(img_sift)
#plt.imshow(img_sift),plt.show()

plt.show()

"""
#FLANN - ORB
FLANN_INDEX_LSH = 6
index_params_o= dict(algorithm = FLANN_INDEX_LSH,
                   table_number = 6, # 12
                   key_size = 12,     # 20
                   multi_probe_level = 1) #2search_params = dict(checks=50)   # or pass empty dictionary

flann_o = cv.FlannBasedMatcher(index_params_o,search_params)

matches_o = flann_o.knnMatch(des1_o, des2_o, k=2)

matchesMask_o = [[0,0] for i in range(len(matches_o))]

#Ratio test (ORB)
for i,(m,n) in enumerate(matches_o):
    if m.distance < 0.72*n.distance:
        matchesMask_o[i]=[1,0]

draw_params_o = dict(matchColor = (0,255,0),
                   singlePointColor = (255,0,0),
                   matchesMask = matchesMask_o,
                   flags = cv.DrawMatchesFlags_DEFAULT)

img_orb = cv.drawMatchesKnn(gray1,kp1_o,gray2,kp2_o,matches_o,None,**draw_params_o)
"""

#plt.subplot(212)
#plt.imshow(img_orb)
#plt.show()

#Display images

"""
cv.imshow('fast',fast_img)
cv.imshow('sift',sift_img)
cv.imshow('original',gray1)
cv.waitKey(0)
cv.destroyAllWindows()
"""


#https://docs.opencv.org/3.4/df/d0c/tutorial_py_fast.html