-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathrun.py
196 lines (163 loc) · 8.21 KB
/
run.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
187
188
189
190
191
192
193
194
195
196
#!/usr/bin/env python2
# -*- coding: utf-8 -*-
"""
Created on Thu Nov 26 04:18:25 2020
@author: jhs
"""
def data_list_load():
file_list = os.listdir('mod_data/landmark')
file_list_int = np.zeros(len(file_list), dtype=int)
for i in range(len(file_list)):
file_list_int[i] = int(file_list[i][0:4])
return file_list_int
def data_load(data_id):
# background = mpimg.imread('../data/background/' + str(data_id) + '.jpg')
landmark = np.genfromtxt('mod_data/landmark/' + str(data_id) +'_landmarks.csv', skip_header=1, delimiter=',',dtype = int)
landmark[:,1] = landmark[:,1] % 10000
recordingMeta = np.genfromtxt('mod_data/recordingMeta/' + str(data_id) + '_recordingMeta.csv', skip_header=1, delimiter = ',')
recordingMeta[3] = recordingMeta[3] % 10000
tracks = np.genfromtxt('mod_data/tracks/' + str(data_id) + '_tracks.csv', skip_header=1, delimiter = ',')
tracksMeta = np.genfromtxt('mod_data/tracksMeta/' + str(data_id) + '_trackMeta.csv', skip_header=1, delimiter = ',')
tracksMeta = np.delete(tracksMeta, -1, -1)
tracksClass = []
with open('mod_data/tracksMeta/' + str(data_id) + '_trackMeta.csv', "r") as tmp_file:
csvReader = csv.reader(tmp_file)
header = next(csvReader)
class_index = header.index("class")
for row in csvReader:
class_tmp = row[class_index]
tracksClass.append(class_tmp)
return landmark, recordingMeta, tracks, tracksMeta, tracksClass
def coordinate_conversion(tracks, landmark, recordingMeta, origin_GT):
global center_GT
global landmark1_GT
global landmark2_GT
global landmark3_GT
global landmark1
global landmark2
global landmark3
meter_per_pixel = recordingMeta[15]
new_tracks = np.zeros_like(tracks)
new_tracks[:] = tracks[:]
landmark1_GT = np.asarray([origin_GT[0]])
landmark2_GT = np.asarray([origin_GT[1]])
landmark3_GT = np.asarray([origin_GT[2]])
center_GT = [(landmark1_GT[0, 0] + landmark2_GT[0, 0] + landmark3_GT[0, 0]) / 3, (landmark1_GT[0, 1] + landmark2_GT[0, 1] + landmark3_GT[0, 1]) / 3]
for i in range(len(landmark)):
print(i)
cur_frame = landmark[i,1]
landmark1 = np.asarray([[landmark[i, 2] * meter_per_pixel, -landmark[i, 3] * meter_per_pixel]])
landmark2 = np.asarray([[landmark[i, 4] * meter_per_pixel, -landmark[i, 5] * meter_per_pixel]])
landmark3 = np.asarray([[landmark[i, 6] * meter_per_pixel, -landmark[i, 7] * meter_per_pixel]])
center = [(landmark1[0, 0] + landmark2[0, 0] + landmark3[0, 0]) / 3, (landmark1[0, 1] + landmark2[0, 1] + landmark3[0, 1]) / 3]
res = minimize(f, [center_GT[0] - center[0], center_GT[1] - center[1], 0], method='Nelder-Mead', tol=1e-10)
trans_x = res.x[0]
trans_y = res.x[1]
rot = res.x[2]
veh_list = np.where(tracks[:,2]==cur_frame)[0]
for j in range(len(veh_list)):
cur_pos = np.asarray([tracks[veh_list[j], 4:6]])
theta_1 = np.rad2deg(np.arctan2(cur_pos[0][1], cur_pos[0][0]))
x_1 = trans_x + np.sqrt(cur_pos[0][0] ** 2 + cur_pos[0][1] ** 2) * np.cos(np.deg2rad(rot + theta_1))
y_1 = trans_y + np.sqrt(cur_pos[0][0] ** 2 + cur_pos[0][1] ** 2) * np.sin(np.deg2rad(rot + theta_1))
new_tracks[veh_list[j], 4:6] = np.asarray([x_1, y_1])
new_tracks[veh_list[j], 6] = new_tracks[veh_list[j], 6] + rot - 90
return new_tracks
def f(x):
trans_x = x[0]
trans_y = x[1]
rot = x[2]
theta_1 = np.rad2deg(np.arctan2(landmark1[0][1], landmark1[0][0]))
theta_2 = np.rad2deg(np.arctan2(landmark2[0][1], landmark2[0][0]))
theta_3 = np.rad2deg(np.arctan2(landmark3[0][1], landmark3[0][0]))
x_1 = trans_x + np.sqrt(landmark1[0][0]**2 + landmark1[0][1]**2) * np.cos(np.deg2rad(rot + theta_1))
y_1 = trans_y + np.sqrt(landmark1[0][0]**2 + landmark1[0][1]**2) * np.sin(np.deg2rad(rot + theta_1))
x_2 = trans_x + np.sqrt(landmark2[0][0] ** 2 + landmark2[0][1] ** 2) * np.cos(np.deg2rad(rot + theta_2))
y_2 = trans_y + np.sqrt(landmark2[0][0] ** 2 + landmark2[0][1] ** 2) * np.sin(np.deg2rad(rot + theta_2))
x_3 = trans_x + np.sqrt(landmark3[0][0] ** 2 + landmark3[0][1] ** 2) * np.cos(np.deg2rad(rot + theta_3))
y_3 = trans_y + np.sqrt(landmark3[0][0] ** 2 + landmark3[0][1] ** 2) * np.sin(np.deg2rad(rot + theta_3))
landmark1_trans = np.asarray([[x_1, y_1]])
landmark2_trans = np.asarray([[x_2, y_2]])
landmark3_trans = np.asarray([[x_3, y_3]])
return np.linalg.norm(landmark1_GT - landmark1_trans) + np.linalg.norm(landmark2_GT - landmark2_trans) + np.linalg.norm(landmark3_GT - landmark3_trans)
import os
import numpy as np
import csv
import sys
import time
sys.path.extend(['/home/jhs/Desktop/data_driven_scenario_gen/'])
import lcm
from lcm_def.morai_tx import xsim_vehicle_global_info
from lcm_def.morai_tx import xsim_ego_info
from scipy.optimize import minimize, rosen, rosen_der
print('Starting KAIST dataset viewer and replayer')
print('Data list loading ...\n')
file_list_int = data_list_load()
print('------------------------------------------------------------')
for i in range(len(file_list_int)):
print('File_id : ' + str(file_list_int[i]), ' File_index : ' + str(i))
print('------------------------------------------------------------')
print('\n')
selected_file_index = input('Select data file index from above :')
while True:
try:
selected_file_index = int(selected_file_index)
if selected_file_index < len(file_list_int) - 1:
break
else:
print('wrong data file index')
selected_file_index = input('Select data file index from above :')
except:
print('wrong data file index')
selected_file_index = input('Select data file index from above :')
selected_scenario_id = file_list_int[selected_file_index]
print('scenario ' + str(selected_scenario_id) + ' is selected')
print('\n')
print('Data loading ....')
landmark, recordingMeta, tracks, tracksMeta, tracksClass = data_load(selected_scenario_id)
origin_GT = [[641.484, -1080.898],
[653.099, -1110.089],
[629.438, -1119.350]]
new_tracks = coordinate_conversion(tracks, landmark, recordingMeta, origin_GT)
init_time = time.time() * 10**9
timer_origin = init_time
timer = 0
fps = 29.97
vehicle_state_lcm = lcm.LCM()
ego_state_lcm = lcm.LCM()
vehicle_state = xsim_vehicle_global_info()
ego_state = xsim_ego_info()
ego_state.x_pos_ego = 0
ego_state.y_pos_ego = 0
ego_state.heading_ego = 0
ego_state.blinker_info = int(0)
ego_state.steering_angle = 0
ego_state.fl_wheel_vel = 0
ego_state.fr_wheel_vel = 0
ego_state.rl_wheel_vel = 0
ego_state.rr_wheel_vel = 0
cur_frame = -1
while True:
timer = time.time() * 10**9 - timer_origin
if timer > 1/fps * 10**9:
cur_frame = cur_frame + 1
timer_origin = time.time() * 10**9
if np.sum(new_tracks[:, 2] == cur_frame) > 0:
vehicle_state.ntime = int(time.time()*10**9 - init_time)
ego_state.ntime = int(time.time()*10**9 - init_time)
vehicle_state.num_of_vehicle = int(np.sum(new_tracks[:, 2] == cur_frame))
vehicle_state.TV_mark = np.zeros(vehicle_state.num_of_vehicle, dtype = int)
vehicle_state.id = new_tracks[new_tracks[:, 2] == cur_frame,1].astype(int)
vehicle_state.x_pos = new_tracks[new_tracks[:, 2] == cur_frame,4]
vehicle_state.y_pos = new_tracks[new_tracks[:, 2] == cur_frame,5]
vehicle_state.x_vel = new_tracks[new_tracks[:, 2] == cur_frame, 9]
vehicle_state.y_vel = new_tracks[new_tracks[:, 2] == cur_frame, 10]
vehicle_state.length = new_tracks[new_tracks[:, 2] == cur_frame, 8]
vehicle_state.width = new_tracks[new_tracks[:, 2] == cur_frame, 7]
vehicle_state.heading = new_tracks[new_tracks[:, 2] == cur_frame, 6]
vehicle_state.lane_id = np.zeros(vehicle_state.num_of_vehicle, dtype = int)
vehicle_state.dist_to_left = np.zeros(vehicle_state.num_of_vehicle)
vehicle_state.dist_to_right = np.zeros(vehicle_state.num_of_vehicle)
vehicle_state_lcm.publish("MORAI_XSIM_VEHICLE_INFO",vehicle_state.encode())
ego_state_lcm.publish("MORAI_EGO_INFO",ego_state.encode())
print('LCM message is published', 'frame : '+str(cur_frame))