depth-estimation-exploration/stereo.py at main · f-wright/depth-estimation-exploration · 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
import numpy as np
import cv2
from matplotlib import pyplot as plt
from matplotlib import cm
import imageio

# # Backpack values
# fx = 7190.247               # lense focal length
# baseline = 174.945          # distance in mm between the two cameras (values from middlebury)
# units = 0.001               # depth units
# doffs=342.523               # x-difference of principal points, following https://vision.middlebury.edu/stereo/data/scenes2014/#description

# texture_threshold = 2000      # 10 by default

# Classroom values
doffs=113.186
baseline=237.604
fx = 3920.793
doffs=113.186

disparities=0
block=23

# # Backpack images
# imgL = cv2.imread('images/im0_left.png', cv2.IMREAD_GRAYSCALE)
# imgR = cv2.imread('images/im0_right.png', cv2.IMREAD_GRAYSCALE)

# Classroom images
imgL = cv2.imread('images/Classroom1-perfect/im0.png', cv2.IMREAD_GRAYSCALE)
imgR = cv2.imread('images/Classroom1-perfect/im1.png', cv2.IMREAD_GRAYSCALE)

plt.imshow(imgL, cmap="gray")
plt.axis('off')
plt.show()

sbm = cv2.StereoBM_create(numDisparities=disparities,blockSize=block)
# sbm.setTextureThreshold(texture_threshold)


# calculate disparities
disparity = sbm.compute(imgL, imgR)
print(disparity)
# show disparity
plt.imshow(disparity)
plt.axis('off')
plt.show()

depth = np.zeros(shape=imgL.shape).astype(float)
depth[disparity > 0] = (fx * baseline) / (doffs + disparity[disparity > 0])

plt.imshow(depth)
plt.show()


# convert from pfm file equation?
pfm = imageio.imread('images/Classroom1-perfect/disp0.pfm')
pfm = np.asarray(pfm)
plt.imshow(pfm)
plt.show()

depth = np.zeros(shape=imgL.shape).astype(float)
depth[pfm > 0] = (fx * baseline) / (doffs + pfm[pfm > 0])
#print(depth)

plt.imshow(depth)
plt.axis('off')
plt.show()