course (#22)

* course * course --------- Co-authored-by: zoudai <2184819232>
OpenHUTB · Jul 7, 2024 · c7d311c · c7d311c
1 parent c52f7da
commit c7d311c
Show file tree

Hide file tree

Showing 11 changed files with 319 additions and 43 deletions.
diff --git a/docs/course/locate_moving.md b/docs/course/locate_moving.md
@@ -0,0 +1,139 @@
+# 车辆按地图选点进行移动
+
+	此示例支持在 **pygame** 窗口对显示区域进行车辆生成并进行移动。
+
+## **环境依赖**
+
+- **python 3.7.9**
+- **carla 0.9.15**
+- **[湖工商场景](https://pan.baidu.com/s/15T1hGoWJ70tVmsTX7-zcSw?pwd=hutb )(WindowsNoEditor)**
+- **[运行 locate_moving.py](../../src/course/locate_moving.py)**
+
+
+
+## CarlaSimulator 类
+
+	该类主要完成设置相机、处理相机图像、将Pygame坐标转换为CARLA世界坐标。
+
+
+
+### 函数初始化__init__
+
+	初始化函数，设置模拟器窗口的宽度和高度，连接CARLA服务器，并获取世界和蓝图库。
+
+```
+def __init__(self, width, height):
+    self.width = width
+    self.height = height
+    self.client = carla.Client('localhost', 2000)
+    self.client.set_timeout(10.0)
+    self.world = self.client.get_world()
+    self.blueprint_library = self.world.get_blueprint_library()
+    self.vehicle_bp = self.blueprint_library.filter('vehicle.*')[0]
+    self.camera = None
+    self.vehicles = []
+    self.fov = VIEW_FOV
+
+```
+
+### setup_camera函数
+
+	设置相机，定义相机的蓝图、分辨率、视场角（FOV），并将相机生成在指定的位置和角度。
+
+```
+def setup_camera(self):
+    camera_bp = self.blueprint_library.find('sensor.camera.rgb')
+    camera_bp.set_attribute('image_size_x', str(self.width))
+    camera_bp.set_attribute('image_size_y', str(self.height))
+    camera_bp.set_attribute('fov', str(self.fov))
+    camera_transform = carla.Transform(carla.Location(x=0, y=0, z=50), carla.Rotation(yaw=270, pitch=-90))
+    self.camera = self.world.spawn_actor(camera_bp, camera_transform)
+
+```
+
+### process_image函数
+
+	处理相机捕获的图像，将其转换为Pygame可显示的格式。
+
+```
+def process_image(self, image):
+    array = np.frombuffer(image.raw_data, dtype=np.uint8)
+    array = array.reshape((self.height, self.width, 4))
+    array = array[:, :, :3]  # Remove alpha channel
+    array = array[:, :, ::-1]  # Convert BGR to RGB
+    surface = pygame.surfarray.make_surface(array.swapaxes(0, 1))
+    return surface
+
+```
+
+### convert_pygame_to_world函数
+
+	将Pygame窗口中的坐标转换为CARLA世界坐标。主要通过相机的视角和Pygame窗口的大小进行计算。
+
+```
+def convert_pygame_to_world(self, x, y):
+    cam_transform = self.camera.get_transform()
+    cam_loc = cam_transform.location
+
+    aspect_ratio = self.width / self.height
+    camera_fov_rad = np.radians(self.fov)
+    camera_fov_y = 2 * np.arctan(np.tan(camera_fov_rad / 2) / aspect_ratio)
+
+    world_units_per_pixel_x = (2 * cam_loc.z * np.tan(camera_fov_rad / 2)) / self.width
+    world_units_per_pixel_y = (2 * cam_loc.z * np.tan(camera_fov_y / 2)) / self.height
+
+    relative_x = (x - self.width / 2) * world_units_per_pixel_x
+    relative_y = (y - self.height / 2) * world_units_per_pixel_y
+
+    world_x = cam_loc.x + relative_x
+    world_y = cam_loc.y - relative_y
+
+    return world_x, world_y
+
+```
+
+### run函数
+
+	运行模拟器，初始化Pygame窗口，设置相机监听器，并处理Pygame事件，如点击事件生成车辆。
+
+```
+def run(self):
+    pygame.init()
+    display = pygame.display.set_mode((self.width, self.height), pygame.HWSURFACE | pygame.DOUBLEBUF)
+    clock = pygame.time.Clock()
+
+    self.setup_camera()
+    self.camera.listen(lambda image: display.blit(self.process_image(image), (0, 0)))
+
+    running = True
+    while running:
+        for event in pygame.event.get():
+            if event.type == pygame.QUIT:
+                running = False
+            elif event.type == pygame.MOUSEBUTTONDOWN:
+                x, y = event.pos
+                world_x, world_y = self.convert_pygame_to_world(x, y)
+
+                spawn_point = carla.Transform(carla.Location(x=world_x, y=-world_y, z=1), carla.Rotation())
+                vehicle = self.world.spawn_actor(self.vehicle_bp, spawn_point)
+                vehicle.set_autopilot()
+                self.vehicles.append(vehicle)
+
+        pygame.display.flip()
+        clock.tick(30)
+
+    for vehicle in self.vehicles:
+        vehicle.destroy()
+    self.camera.destroy()
+    pygame.quit()
+
+```
+
+
+
+![](../img/traffic_course_img/locate_moving.gif)
+
+
+
+
+
diff --git a/docs/course/microscopic_Intersection_demo.md b/docs/course/microscopic_Intersection_demo.md
@@ -4,13 +4,19 @@
 
 此脚本连接到CARLA仿真环境，并定期切换湖工商附近的四个预设路口视角。通过不断改变观众（Spectator）的视角，用户可以观察到不同路口的交通状况。
 
-#### **环境要求**
-
-- Python 3.6 及以上版本
-
-- CARLA 仿真环境
-
-- 安装必要的Python库：
+- #### **环境要求**
+
+  - Python 3.6 及以上版本
+
+  - CARLA 仿真环境
+
+  - 安装必要的Python库：
+
+  - **[spectator.py](https://github.com/OpenHUTB/carla_doc/blob/master/src/course/spectator/spectator.py)**
+
+  - [**湖工商场景**](https://pan.baidu.com/s/15T1hGoWJ70tVmsTX7-zcSw?pwd=hutb )**(WindowsNoEditor)**，并运行**[generate_traffic.py](https://github.com/OpenHUTB/carla_doc/blob/master/src/examples/generate_traffic.py)**
+
+- 安装必要的python库
 
   ```
   pip install carla

diff --git a/docs/course/multi-view_ camera.md b/docs/course/multi-view_ camera.md
@@ -4,11 +4,13 @@
 
 此脚本连接到CARLA仿真环境，在车辆的左右两侧设置相机，并使用Pygame显示相机捕获的实时图像。系统采用多进程架构，确保左右相机图像分别在不同窗口中实时显示。
 
-#### **环境要求**
+- #### **环境要求**
 
-- Python 3.6 及以上版本
-- CARLA 仿真环境
-- 安装必要的Python库：
+  - Python 3.6 及以上版本
+  - CARLA 仿真环境
+  - 安装必要的Python库：
+  - [**camera.py**](https://github.com/OpenHUTB/carla_doc/blob/master/src/course/camera.py)
+  - [**湖工商场景**](https://pan.baidu.com/s/15T1hGoWJ70tVmsTX7-zcSw?pwd=hutb )**(WindowsNoEditor)**
 
 ```
 pip install carla pygame numpy opencv-python

diff --git a/docs/course/navigation.md b/docs/course/navigation.md
@@ -2,11 +2,13 @@
 
 	车辆运动支持定义车辆运动起点和目标位置，为每辆车指定随机速度值，使其沿轨迹移动时不会发生碰撞或模拟碰撞，看到车辆沿指定轨迹行驶以到达其目标位置。
 
-	本示例设计了一个**automatic_control_revised.py**脚本进行实现
+	本示例设计了一个  [**automatic_control_revised.py**](https://github.com/OpenHUTB/carla_doc/blob/master/src/course/navigation/automatic_control_revised.py)  脚本进行实现
 
 ![](../img/traffic_course_img/4.gif)
 
-**注意**  需要将**agents**文件夹位置添加到代码环境变量中，或者复制该文件夹使其与本示例脚本同级。
+**注意**  需要将 [**agents**](https://github.com/OpenHUTB/carla_doc/tree/master/src/carla_agent) 文件夹位置添加到代码环境变量中，或者复制该文件夹使其与本示例脚本同级。
+
+   [**湖工商场景**](https://pan.baidu.com/s/15T1hGoWJ70tVmsTX7-zcSw?pwd=hutb )**(WindowsNoEditor)**
 
 
 
@@ -28,18 +30,20 @@ diming = {
 	定义命令行参数
 
 ```python
-argparser = argparse.ArgumentParser(
-    description=__doc__)
-argparser.add_argument(
-    '--s',
-    default="北门",
-    type=str)
-argparser.add_argument(
-    '--e',
-    default="西门",
-    type=str)
-args = argparser.parse_args()
-
+ argparser = argparse.ArgumentParser(
+        description=__doc__)
+    argparser.add_argument(
+        '--s',
+        default="北门",
+        type=str)
+    argparser.add_argument(
+        '--e',
+        default="西门",
+        type=str)
+    argparser.add_argument(
+        '--speed',
+        default=5,
+        type=int)
 ```
 
 	获取仿真时间对象及设置
@@ -76,9 +80,11 @@ vehicle = world.spawn_actor(ego_vehicle_bp, spawn_point)
 world.tick()
 ```
 
-	生成终点，给起点和终点生成一个车辆运行轨迹
+	生成终点，给起点和终点生成一个车辆运行轨迹，定义初始速度
 
 ```python
+ # 设置初始速度
+ vehicle.set_target_velocity(carla.Vector3D(args.speed, 0, 0))
 # 创建代理
 agent = BehaviorAgent(vehicle, behavior='normal')
 destination = carla.Location(locations[args.e])
@@ -167,4 +173,6 @@ while True:
     # agent.get_local_planner().set_speed(50)
     control = agent.run_step(debug=True)
     vehicle.apply_control(control)
-```
+```
+
+![](../img/traffic_course_img/5.png)
diff --git a/docs/course/object_detection.md b/docs/course/object_detection.md
@@ -1,6 +1,8 @@
 # 基于Carla仿真环境的目标检测
 
-	**object_dection.py**  脚本主要是在 **Carla** 中使用 **YOLOv3** 检测车辆。[相关链接](https://openhutb.github.io/carla_doc/algorithms/perception/)
+	[**object_dection.py**](https://github.com/OpenHUTB/carla_doc/blob/master/src/course/object_detection/object_detection.py)  脚本主要是在 **Carla** 中使用 **YOLOv3** 检测车辆。[相关链接](https://openhutb.github.io/carla_doc/algorithms/perception/)  
+
+     [**湖工商场景**](https://pan.baidu.com/s/15T1hGoWJ70tVmsTX7-zcSw?pwd=hutb )**(WindowsNoEditor)**，并运行**[generate_traffic.py](https://github.com/OpenHUTB/carla_doc/blob/master/src/examples/generate_traffic.py)**
 
 ![](../img/traffic_course_img/3.gif)
 

diff --git a/docs/course/traffic_indicators.md b/docs/course/traffic_indicators.md
@@ -1,8 +1,8 @@
 # 路口评价指标分析计算(HutbCarlaCity)
 
-	为支持高保正的十字路口三维建模，对路口交通流量、路口车均延误、路口饱和、排队长度四种路口真实性评价指标分析。
+	为支持  [**湖工商场景**](https://pan.baidu.com/s/15T1hGoWJ70tVmsTX7-zcSw?pwd=hutb ) **(WindowsNoEditor)** 高保正的十字路口三维建模，设计  **[traffic_indicators.py](https://github.com/OpenHUTB/carla_doc/blob/master/src/course/traffic%20indicators.py)**  脚本对路口交通流量、路口车均延误、路口饱和、排队长度四种路口真实性评价指标分析。
 
-**首先需要对场景中添加车辆，运行generate_traffic.py脚本，生成足够的车辆**
+**首先需要对场景中添加车辆，运行[generate_traffic.py](https://github.com/OpenHUTB/carla_doc/blob/master/src/examples/generate_traffic.py)脚本，生成足够的车辆**
 
 
 

diff --git a/docs/course/trajectory_tracking.md b/docs/course/trajectory_tracking.md
@@ -1,8 +1,8 @@
 # 轨迹跟踪  Drive.py
 
-	本示例使用  **Carla**  模拟器来执行轨迹跟踪任务，主要功能包括连接CARLA服务器、设置模拟环境、生成车辆、读取和处理路径点数据、控制车辆运动、记录和保存仿真结果下面对于主要函数进行介绍。
-
+	本示例使用  **Carla**  模拟器来执行轨迹跟踪任务，  [**Drive.py**](https://github.com/OpenHUTB/carla_doc/blob/master/src/course/trajectory_tracking/Drive.py)  主要功能包括连接CARLA服务器、设置模拟环境、生成车辆、读取和处理路径点数据、控制车辆运动、记录和保存仿真结果下面对于主要函数进行介绍。
 
+  基于 [**湖工商场景**](https://pan.baidu.com/s/15T1hGoWJ70tVmsTX7-zcSw?pwd=hutb )**(WindowsNoEditor)**，首先需要运行[generate_traffic.py](https://github.com/OpenHUTB/carla_doc/blob/master/src/examples/generate_traffic.py)
 
 ## 定义CARLA设置
 

diff --git a/docs/img/traffic_course_img/5.png b/docs/img/traffic_course_img/5.png
diff --git a/docs/img/traffic_course_img/locate_moving.gif b/docs/img/traffic_course_img/locate_moving.gif
diff --git a/src/course/locate_moving.py b/src/course/locate_moving.py
@@ -0,0 +1,107 @@
+import carla
+import pygame
+import numpy as np
+import random
+
+# 设置Pygame窗口大小
+VIEW_WIDTH = 1920 // 2
+VIEW_HEIGHT = 1080 // 2
+VIEW_FOV = 90
+
+
+class CarlaSimulator:
+    def __init__(self, width, height):
+        self.width = width
+        self.height = height
+        self.client = carla.Client('localhost', 2000)
+        self.client.set_timeout(10.0)
+        self.world = self.client.get_world()
+        self.blueprint_library = self.world.get_blueprint_library()
+        self.vehicle_bp = self.blueprint_library.filter('vehicle.*')[0]
+        self.camera = None
+        self.vehicles = []
+        self.fov = VIEW_FOV
+
+    def setup_camera(self):
+        camera_bp = self.blueprint_library.find('sensor.camera.rgb')
+        camera_bp.set_attribute('image_size_x', str(self.width))
+        camera_bp.set_attribute('image_size_y', str(self.height))
+        camera_bp.set_attribute('fov', str(self.fov))
+        camera_transform = carla.Transform(carla.Location(x=0, y=0, z=50), carla.Rotation(yaw=270, pitch=-90))
+        self.camera = self.world.spawn_actor(camera_bp, camera_transform)
+
+    def process_image(self, image):
+        array = np.frombuffer(image.raw_data, dtype=np.uint8)
+        array = array.reshape((self.height, self.width, 4))
+        array = array[:, :, :3]  # Remove alpha channel
+        array = array[:, :, ::-1]  # Convert BGR to RGB
+        surface = pygame.surfarray.make_surface(array.swapaxes(0, 1))
+        return surface
+
+    def convert_pygame_to_world(self, x, y):
+        # 获取相机位置和旋转
+        cam_transform = self.camera.get_transform()
+        cam_loc = cam_transform.location
+
+        # 计算相机的水平和垂直视角
+        aspect_ratio = self.width / self.height
+        camera_fov_rad = np.radians(self.fov)
+        camera_fov_y = 2 * np.arctan(np.tan(camera_fov_rad / 2) / aspect_ratio)
+
+        # 计算每个像素在实际世界中的对应距离
+        world_units_per_pixel_x = (2 * cam_loc.z * np.tan(camera_fov_rad / 2)) / self.width
+        world_units_per_pixel_y = (2 * cam_loc.z * np.tan(camera_fov_y / 2)) / self.height
+
+        # 将Pygame坐标转换为相对于图像中心的坐标
+        relative_x = (x - self.width / 2) * world_units_per_pixel_x
+        relative_y = (y - self.height / 2) * world_units_per_pixel_y
+
+        # 转换为世界坐标
+        world_x = cam_loc.x + relative_x
+        world_y = cam_loc.y - relative_y  # 注意这里是减去，因为y轴方向是相反的
+
+        return world_x, world_y
+
+    def run(self):
+        pygame.init()
+        display = pygame.display.set_mode((self.width, self.height), pygame.HWSURFACE | pygame.DOUBLEBUF)
+        clock = pygame.time.Clock()
+
+        self.setup_camera()
+        self.camera.listen(lambda image: display.blit(self.process_image(image), (0, 0)))
+
+        spectator = self.world.get_spectator()
+
+        running = True
+        while running:
+            for event in pygame.event.get():
+                if event.type == pygame.QUIT:
+                    running = False
+                elif event.type == pygame.MOUSEBUTTONDOWN:
+                    x, y = event.pos
+                    # 将Pygame坐标转换为CARLA世界坐标
+                    world_x, world_y = self.convert_pygame_to_world(x, y)
+                    # print(f'World coordinates: ({world_x}, {world_y})')
+
+                    # self.world.debug.draw_string(carla.Location(x=world_x, y=-world_y, z=1), 'ss', life_time=1000,
+                    #                              color=carla.Color(255, 0, 0))
+                    spawn_point = carla.Transform(carla.Location(x=world_x, y=-world_y, z=1), carla.Rotation())
+                    vehicle = self.world.spawn_actor(self.vehicle_bp, spawn_point)
+                    vehicle.set_autopilot()
+                    # if self.vehicles is not None:
+                    #     last_vehicle_transform = self.vehicles[-1].get_transform()
+                    #     spectator.set_transform(last_vehicle_transform)
+                    self.vehicles.append(vehicle)
+
+            pygame.display.flip()
+            clock.tick(30)
+
+        for vehicle in self.vehicles:
+            vehicle.destroy()
+        self.camera.destroy()
+        pygame.quit()
+
+
+if __name__ == '__main__':
+    simulator = CarlaSimulator(VIEW_WIDTH, VIEW_HEIGHT)
+    simulator.run()