forked from Thomahawkuru/Realsense-Surround-Tracking
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Merger.py
186 lines (150 loc) · 7.92 KB
/
Merger.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
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
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
import time
import numpy as np
from MultiDetector import MultiDetector # Assuming your class is saved in Python_Realsense_Multi.py
import threading
import cv2
from mpl_toolkits.mplot3d import Axes3D
import matplotlib.pyplot as plt
from matplotlib.lines import Line2D
class Merger:
def __init__(self, camera_config_path, detection_type='mask', threshold=0.1, show=True, plot=True):
self.camera_config_path = camera_config_path
self.detection_type = detection_type
self.threshold = threshold
self.show = show
self.plot = plot
self.detector = MultiDetector(camera_config_path=self.camera_config_path, detection_type=self.detection_type, show=True, draw=True, plot=False, verbose=False)
if plot:
self._init_plot()
def _init_plot(self):
"""Initialize 2D plotting with matplotlib."""
self.fig, self.ax = plt.subplots(figsize=(10, 10)) # 2D plot
self.ax.set_xlabel('X') # X-axis
self.ax.set_ylabel('Z') # Z on the Y-axis
self.ax.set_title('Top Down view of detections, single and merged')
# Set axis limits
self.ax.set_xlim([-10, 10])
self.ax.set_ylim([-10, 10]) # Z-axis limits now on Y-axis
# Add an arrow in the positive Z direction at (0, 0)
self.ax.annotate('', xy=(0, 0), xytext=(0, -0.3),
arrowprops=dict(facecolor='green', shrink=0.05, width=2, headwidth=10))
# Initialize scatter objects for single and merged detections
self.single_scatter = self.ax.scatter([], [], c='b', marker='o', label='Single')
self.merged_scatter = self.ax.scatter([], [], c='r', marker='o', label='Merged')
# Create a custom arrow legend entry using Line2D
arrow_legend = Line2D([0], [0], color='green', marker='|', linestyle='None', markersize=10, label='Robot')
# Add the arrow to the legend along with scatter plots
self.ax.legend(handles=[self.single_scatter, self.merged_scatter, arrow_legend], loc='upper right')
def calculate_average_position(self, detections):
"""Calculate the average position of detections."""
avg_x = np.mean([d[0] for d in detections])
avg_y = np.mean([d[1] for d in detections])
avg_z = np.mean([d[2] for d in detections])
return [avg_x, avg_y, avg_z]
def find_similar_detections(self, detection_list):
"""Group detections by object names and then compare positions across different cameras."""
name_groups = {}
for detection in detection_list:
obj_x, obj_y, obj_z, object_ids, object_names, cam_idx = detection
for name, obj_id, x, y, z in zip(object_names, object_ids, obj_x, obj_y, obj_z):
if name not in name_groups:
name_groups[name] = {
'cam_idx': [cam_idx],
'object_id': [obj_id],
'Position': [[x, y, z]],
}
else:
name_groups[name]['cam_idx'].append(cam_idx)
name_groups[name]['object_id'].append(obj_id)
name_groups[name]['Position'].append([x, y, z])
similar_detections = []
for name, data in name_groups.items():
positions = data['Position']
cam_indices = data['cam_idx']
object_ids = data['object_id']
group = []
seen = set()
for i in range(len(object_ids)):
if object_ids[i] in seen:
continue
current_position = np.array(positions[i])
current_cam_idx = cam_indices[i]
current_obj_id = object_ids[i]
similar_group = {
'object_name': name,
'object_id': [current_obj_id],
'positions': [current_position],
'cameras': [current_cam_idx]
}
seen.add(object_ids[i])
for j in range(i + 1, len(positions)):
if object_ids[j] in seen:
continue
other_position = np.array(positions[j])
other_cam_idx = cam_indices[j]
other_obj_id = object_ids[j]
distance = np.linalg.norm(current_position - other_position)
if distance < self.threshold and current_cam_idx != other_cam_idx:
similar_group['object_id'].append(other_obj_id)
similar_group['positions'].append(other_position)
similar_group['cameras'].append(other_cam_idx)
seen.add(object_ids[j])
if len(similar_group['positions']) > 1:
avg_position = self.calculate_average_position(similar_group['positions'])
similar_group['average_position'] = avg_position
group.append(similar_group)
similar_detections.extend(group)
return similar_detections
def plot_2d_results(self, similar_detections):
"""Plot the raw detections and the merged results in a 2D plot."""
# Separate x, z coordinates for raw and merged positions (since Z is now on the Y-axis)
pos_x, pos_z = [], []
avg_x, avg_z = [], []
# Organize detection data into the correct format for plotting
for det in similar_detections:
if len(det['object_id']) > 1: # Merged detections
avg_pos = det['average_position']
avg_x.append(float(avg_pos[0])) # X
avg_z.append(float(avg_pos[2])) # Z (on Y-axis)
else: # Single detections
pos = det['positions'][0]
pos_x.append(pos[0]) # X
pos_z.append(pos[2]) # Z (on Y-axis)
# Update scatter plot data
self.single_scatter.set_offsets(np.c_[pos_x, pos_z])
self.merged_scatter.set_offsets(np.c_[avg_x, avg_z])
# Redraw the plot
plt.draw()
plt.pause(0.001) # This is required to keep the plot interactive
def merge_and_plot_detections(self):
"""Merge detections and plot 3D results."""
self.detector.start()
try:
while True:
combined_detections = []
for idx, detection in enumerate(self.detector.detection_list):
if detection is not None:
combined_detections.append(detection)
similar_detections = self.find_similar_detections(combined_detections)
# Print merged detections with similar names and positions within the threshold
print("\n--- Merged Detections (within threshold) ---")
if not similar_detections:
print("No merged detections found within the threshold.")
for det in similar_detections:
if len(det['object_id']) > 1:
avg_pos = det['average_position']
print(f"Detected duplicate {det['object_name']} (ID: {det['object_id']}) at average position {avg_pos} from cameras {det['cameras']}")
else:
pos = det['positions'][0]
print(f"Detected {det['object_name']} (ID: {det['object_id']}) at position {pos} from camera {det['cameras']}")
if self. plot: self.plot_2d_results(similar_detections)
if self.show: self.detector.show_combined_frames()
time.sleep(0.1)
if cv2.waitKey(1) & 0xFF == ord("q"):
break
finally:
plt.close()
self.detector.stop()
if __name__ == "__main__":
merger = Merger(camera_config_path='CAMERAS_Jackal04.json', threshold=0.35, show=True, plot=True)
merger.merge_and_plot_detections()