forked from ThisisWilli/binocular_vision
/
SGBM_matching.py
95 lines (82 loc) · 4.43 KB
/
SGBM_matching.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
'''
@project : binocular_vision
@author : Hoodie_Willi
#@description: $将两张图片进行SGBM匹配，得出深度图, 计算深度
#@time : 2019-04-18 21:39:34
'''
import numpy as np
import cv2
import time
import camera_configs
'''
创建鼠标点击事件，返回三维坐标
'''
def callbackFunc(e, x, y, f, p):
if e == cv2.EVENT_LBUTTONDOWN:
print(threeD[y][x])
'''
改变参数重新计算视差
'''
def SGBM_update(val=0):
global threeD
global SGBM_num
global SGBM_blockSize
SGBM_blockSize = cv2.getTrackbarPos('blockSize', 'SGBM_disparity') # 从滑动条中获取值
if SGBM_blockSize % 2 == 0:
SGBM_blockSize += 1
if SGBM_blockSize < 5:
SGBM_blockSize = 5
SGBM_stereo.setBlockSize(SGBM_blockSize)
SGBM_num = cv2.getTrackbarPos('num_disp', 'SGBM_disparity')
num_disp = SGBM_num * 16
SGBM_stereo.setNumDisparities(num_disp)
SGBM_stereo.setUniquenessRatio(cv2.getTrackbarPos('unique_Ratio', 'SGBM_disparity'))
SGBM_stereo.setSpeckleWindowSize(cv2.getTrackbarPos('spec_WinSize', 'SGBM_disparity'))
SGBM_stereo.setSpeckleRange(cv2.getTrackbarPos('spec_Range', 'SGBM_disparity'))
SGBM_stereo.setDisp12MaxDiff(cv2.getTrackbarPos('disp12MaxDiff', 'SGBM_disparity'))
print('computing SGBM_disparity...')
disp = SGBM_stereo.compute(imgL, imgR).astype(np.float32) / 16.0
threeD = cv2.reprojectImageTo3D(disp, camera_configs.Q)
# cv2.imshow('left', imgL)
# cv2.imshow('right', imgR)
cv2.imshow('SGBM_disparity', (disp - min_disp) / num_disp)
cv2.imwrite("./detect_img/SGBM_depth.jpg", (disp - min_disp) / num_disp)
if __name__ == "__main__":
start = time.clock()
SGBM_blockSize = 5 # 一个匹配块的大小,大于1的奇数
SGBM_num = 2
min_disp = 0 # 最小的视差值，通常情况下为0
num_disp = SGBM_num * 16 # 192 - min_disp #视差范围，即最大视差值和最小视差值之差，必须是16的倍数。
# blockSize = blockSize #匹配块大小（SADWindowSize），必须是大于等于1的奇数，一般为3~11
uniquenessRatio = 6 # 视差唯一性百分比， 视差窗口范围内最低代价是次低代价的(1 + uniquenessRatio/100)倍时，最低代价对应的视差值才是该像素点的视差，否则该像素点的视差为 0，通常为5~15.
speckleRange = 2 # 视差变化阈值，每个连接组件内的最大视差变化。如果你做斑点过滤，将参数设置为正值，它将被隐式乘以16.通常，1或2就足够好了
speckleWindowSize = 60 # 平滑视差区域的最大尺寸，以考虑其噪声斑点和无效。将其设置为0可禁用斑点过滤。否则，将其设置在50-200的范围内。
disp12MaxDiff = 200 # 左右视差图的最大容许差异（超过将被清零），默认为 -1，即不执行左右视差检查。
P1 = 600 # 惩罚系数，一般：P1=8*通道数*SADWindowSize*SADWindowSize，P2=4*P1
P2 = 2400 # p1控制视差平滑度，p2值越大，差异越平滑
imgL = cv2.imread('./detect_img/left.jpg')
imgR = cv2.imread('./detect_img/right.jpg')
cv2.namedWindow('SGBM_disparity')
cv2.setMouseCallback('SGBM_disparity', callbackFunc, None)
cv2.createTrackbar('blockSize', 'SGBM_disparity', SGBM_blockSize, 21, SGBM_update)
cv2.createTrackbar('num_disp', 'SGBM_disparity', SGBM_num, 20, SGBM_update)
cv2.createTrackbar('spec_Range', 'SGBM_disparity', speckleRange, 50, SGBM_update) # 设置trackbar来调节参数
cv2.createTrackbar('spec_WinSize', 'SGBM_disparity', speckleWindowSize, 200, SGBM_update)
cv2.createTrackbar('unique_Ratio', 'SGBM_disparity', uniquenessRatio, 50, SGBM_update)
cv2.createTrackbar('disp12MaxDiff', 'SGBM_disparity', disp12MaxDiff, 250, SGBM_update)
SGBM_stereo = cv2.StereoSGBM_create(
minDisparity=min_disp, # 最小的视差值
numDisparities=num_disp, # 视差范围
blockSize=SGBM_blockSize, # 匹配块大小（SADWindowSize）
uniquenessRatio=uniquenessRatio, # 视差唯一性百分比
speckleRange=speckleRange, # 视差变化阈值
speckleWindowSize=speckleWindowSize,
disp12MaxDiff=disp12MaxDiff, # 左右视差图的最大容许差异
P1=P1, # 惩罚系数
P2=P2
)
SGBM_update()
end = time.clock()
print('SGBM Running time: %s Seconds' % (end - start))
cv2.waitKey(0) # 等待键盘输入，如果输入为0则无限等待
cv2.destroyAllWindows() # 关闭所有地窗口