#!/usr/bin/env python
# Copyright (c) 2019 Computer Vision Center (CVC) at the Universitat Autonoma de ?
# Barcelona (UAB).
#
# This work is licensed under the terms of the MIT license.
# For a copy, see <https://opensource.org/licenses/MIT>.
import glob
import os
import sys
import numpy as np
import cv2
try:
sys.path.append(glob.glob('../carla/dist/carla-*%d.%d-%s.egg' % (
sys.version_info.major,
sys.version_info.minor,
'win-amd64' if os.name == 'nt' else 'linux-x86_64'))[0])
## Find the version of python started
except IndexError:
pass
import carla
import random
import time
IM_WIDTH = 640
IM_HEIGHT = 480
SHOW_PREVIEW = False
actor_list = []
def process_img(image, window_name):
i = np.array(image.raw_data)
# print(i. shape)
i2 = i.reshape((IM_HEIGHT, IM_WIDTH, 4))
i3 = i2[:, :, :3]
cv2.imshow(window_name, i3)
cv2.waitKey(1)
return i3/255.0
# Define a callback function to process camera images
def process_image(image, window_name):
# Convert the image to a numpy array
array = np.frombuffer(image.raw_data, dtype=np.uint8).reshape((image.height, image.width, 4))
# Remove the alpha channel
array = array[:, :, :3]
# Flip the image vertically (OpenCV uses a different coordinate system than Carla)
array = np. flipud(array)
# Show the image in the window
cv2.imshow(window_name, array)
def collision_data(self, event):
self.collision_hist.append(event)
def add_camera_to_vehicle(vehicle, spawn_point, im_width=640, im_height=480):
"""
Add a camera to the vehicle
:param vehicle: vehicle object
:param spawn_point: The position where the camera spawns
:param im_width: The width of the camera image, the default value is 640
:param im_height: The height of the camera image, the default value is 480
:return: return camera object
"""
blueprint_library = vehicle. get_world(). get_blueprint_library()
cam_bp = blueprint_library. find("sensor. camera. rgb")
cam_bp.set_attribute("image_size_x", f"{im_width}")
cam_bp.set_attribute("image_size_y", f"{im_height}")
cam_bp.set_attribute("fov", "110")
camera = vehicle.get_world().spawn_actor(cam_bp, spawn_point, attach_to=vehicle)
return camera
def add_camera_to_vehicle(vehicle, spawn_point, im_width=640, im_height=480):
"""
Add a camera to the vehicle
:param vehicle: vehicle object
:param spawn_point: The position where the camera spawns
:param im_width: The width of the camera image, the default value is 640
:param im_height: The height of the camera image, the default value is 480
:return: return camera object
"""
blueprint_library = vehicle. get_world(). get_blueprint_library()
cam_bp = blueprint_library. find("sensor. camera. rgb")
cam_bp.set_attribute("image_size_x", f"{im_width}")
cam_bp.set_attribute("image_size_y", f"{im_height}")
cam_bp.set_attribute("fov", "110")
camera = vehicle.get_world().spawn_actor(cam_bp, spawn_point, attach_to=vehicle)
return camera
def add_collision_sensor_to_vehicle(vehicle):
"""
Add a crash sensor to the vehicle
:param vehicle: vehicle object
:return: Returns the collision sensor object
"""
blueprint_library = vehicle. get_world(). get_blueprint_library()
collision_bp = blueprint_library.find("sensor.other.collision")
collision_sensor = vehicle.get_world().spawn_actor(collision_bp, carla.Transform(), attach_to=vehicle)
return collision_sensor
def process_lidar_measurement(measurement):
points = np.frombuffer(measurement.raw_data, dtype=np.dtype('f4'))
points = np. reshape(points, (int(points. shape[0] / 4), 4))
lidar_data = np.array(points[:, :3])
# Display the point cloud image
image = np.zeros((600, 800, 3), dtype=np.uint8)
lidar_data *= 50
lidar_data += [400, 300, 0]
lidar_data = lidar_data.astype(np.int32)
for point in lidar_data:
cv2. circle(image, (point[0], point[1]), 1, (0, 0, 255), -1)
cv2.imshow('Lidar data', image)
cv2.waitKey(1)
def add_lidar_to_vehicle(vehicle, spawn_point, num_points=100000):
# define radar parameters
blueprint_library = vehicle. get_world(). get_blueprint_library()
lidar_bp = blueprint_library. find("sensor. lidar. ray_cast")
lidar_bp.set_attribute("channels", "360") # 360°
lidar_bp.set_attribute("range", "500")
lidar_bp.set_attribute("points_per_second", "100000")
lidar_bp.set_attribute("rotation_frequency", "200") # set rotation_frequency to 40
lidar_bp.set_attribute("upper_fov", "50") # set upper_fov to 30
lidar_bp.set_attribute("lower_fov", "-45")
lidar = vehicle.get_world().spawn_actor(lidar_bp, spawn_point, attach_to=vehicle)
return lidar
def start():
client = carla. Client("localhost", 2000)
client.set_timeout(2.2)
world = client. get_world()
world = client.load_world('Town05')
blueprint_library = world. get_blueprint_library()
bp = blueprint_library. filter("model3")[0]
print(bp)
spawn_point = random.choice(world.get_map().get_spawn_points())
vehicle = world.spawn_actor(bp, spawn_point)
# vehicle. set_autopilot(True)
vehicle.apply_control(carla.VehicleControl(throttle=1.0, steer=0.0))
actor_list.append(vehicle)
spawn_point_1 = carla. Transform(carla. Location(x=1.5, y=0, z=2.4))
camera_1 = add_camera_to_vehicle(vehicle, spawn_point_1)
spawn_point_2 = carla. Transform(carla. Location(x=-1.5, y=0, z=2.4))
camera_2 = add_camera_to_vehicle(vehicle, spawn_point_2)
spawn_point_3 = carla. Transform(carla. Location(x=0, y=0, z=2.5))
lidar_1 = add_lidar_to_vehicle(vehicle, spawn_point_3)
# configure and start radar
actor_list.append(lidar_1)
col = add_collision_sensor_to_vehicle(vehicle)
actor_list.append(camera_1)#### is very important for displaying images#####
actor_list.append(camera_2) ####It is very important for displaying images#####
actor_list.append(col)
camera_1.listen(lambda data: process_img(data, "camera_1"))
camera_2.listen(lambda data: process_img(data, "camera_2"))
lidar_1.listen(process_lidar_measurement)
col.listen(lambda event: collision_data(event))
try:
start()
time. sleep(100)
finally:
for actor in actor_list:
actor. destroy()
print("ALL cleaned up!!!")
The knowledge points of the article match the official knowledge files, and you can further learn relevant knowledgePython entry skill treeHomepageOverview 258581 people are studying systematically