vision_physics_bridge/plot_utils.py at main · DAIRLab/vision_physics_bridge · 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
115
116
117
118
119
120
121
122
123
import numpy as np
import scipy.spatial as sp
import matplotlib.pyplot as plt

from math_utils import camera_to_world, transform_to_camera

GT_POSE_DIR = "/home/cnets-vision/mengti_ws/robot_filter/tagslam_poses/"
OUTPUT_POSE_DIR = "/home/cnets-vision/mengti_ws/poses/"

def get_cosine_sim(frame_id):
    """Compare the output of BundleTrack with the ground-truth poses of tagslam.
    """
    gt_pose = np.loadtxt(GT_POSE_DIR+'%04i.txt'% frame_id)
    output_pose = np.loadtxt(OUTPUT_POSE_DIR+'%04i.txt'% frame_id)
    return 1 - sp.distance.cdist(gt_pose, output_pose, 'cosine')

def get_angle(P, Q):
    R = np.dot(P, Q.T)
    theta = (np.trace(R) -1)/2
    return np.arccos(theta) * (180/np.pi)

def plot_xyz(start_frame, end_frame):
    """Plot the x, y, z of BundleTrack output versus ground-truth poses of tagslam.
    """
    output_x, output_y, output_z = [], [], [] #translation
    gt_x, gt_y, gt_z = [], [], []
    angles = []

    for frame_id in range(start_frame, end_frame):
        gt_frame = frame_id+700
        gt_pose = np.loadtxt(GT_POSE_DIR+'%04i.txt'% gt_frame)
        output_pose = np.loadtxt(OUTPUT_POSE_DIR+'%04i.txt'% frame_id)
        # output_pose = transform_to_camera(output_pose)
        output_pose = camera_to_world(output_pose)
        angle = get_angle(output_pose[:3, :3], gt_pose[:3, :3])
        angles.append(angle)
        output_x.append(output_pose[0, 3])
        output_y.append(output_pose[1, 3])
        output_z.append(output_pose[2, 3])

        gt_x.append(gt_pose[0, 3])
        gt_y.append(gt_pose[1, 3])
        gt_z.append(gt_pose[2, 3])
    output_x, output_y, output_z = np.array(output_x), np.array(output_y), np.array(output_z)
    gt_x, gt_y, gt_z = np.array(gt_x), np.array(gt_y), np.array(gt_z)

    x = np.arange(0, output_x.shape[0])
    plt.plot(x, output_x, label="Bundletrack")
    plt.plot(x, gt_x, label="ground-truth")
    plt.xlabel("X-axis")
    plt.ylabel("Y-axis")
    plt.title("Position x")
    plt.legend()
    plt.show()

    plt.plot(x, output_y, label="BundleTrack")
    plt.plot(x, gt_y, label="ground-truth")
    plt.xlabel("X-axis")
    plt.ylabel("Y-axis")
    plt.title("Position y")
    plt.legend()
    plt.show()

    plt.plot(x, output_z, label="BundleTrack")
    plt.plot(x, gt_z, label="ground-truth")
    plt.xlabel("X-axis")
    plt.ylabel("Y-axis")
    plt.title("Position z")
    plt.legend()
    plt.show()

    plt.plot(x, angles)
    plt.xlabel("X-axis")
    plt.ylabel("angle diff (degrees)")
    plt.title("Angle difference")
    plt.show()

def plot_ground_truth(start_frame, end_frame):
    gt_x, gt_y, gt_z = [], [], []
    for frame_id in range(start_frame, end_frame):
        gt_pose = np.loadtxt(GT_POSE_DIR+'%04i.txt'% frame_id)
        gt_x.append(gt_pose[0, 3])
        gt_y.append(gt_pose[1, 3])
        gt_z.append(gt_pose[2, 3])

    gt_x, gt_y, gt_z = np.array(gt_x), np.array(gt_y), np.array(gt_z)
    x = np.arange(0, gt_x.shape[0])
    plt.plot(x, gt_x, label="ground-truth")
    plt.xlabel("Frame")
    plt.ylabel("X axis")
    plt.title("Position x")
    plt.legend()
    plt.show()

    plt.plot(x, gt_y, label="ground-truth")
    plt.xlabel("Frame")
    plt.ylabel("Y axis")
    plt.title("Position y")
    plt.legend()
    plt.show()

    plt.plot(x, gt_z, label="ground-truth")
    plt.xlabel("Frame")
    plt.ylabel("Z axis")
    plt.title("Position z")
    plt.legend()
    plt.show()


if __name__ ==  '__main__':
    plot_xyz(1, 950)
    # plot_ground_truth(701, 4396)


    # from PIL import Image
    # import imageio
    # img_dir = "./depth_data/0001.png"
    # # img_dir= "/home/cnets-vision/mengti_ws/BundleTrack/Data/YCBINEOAT/bleach0/depth/1578966042136484471.png"
    # # image = Image.open(img_dir)
    # # pixels = list(image.getdata())
    # image = imageio.imread(img_dir)
    # print(image[100][100])
    # print(image.dtype)