|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/ros2-self-driving-car-with-deep-learning-and-computer-vision/
课程评论：没有评论
**课程名称：** ROS2 自动驾驶汽车：深度学习与计算机视觉 **课程概述：** 本课程将指导您使用ROS2从头开始构建一辆基于RGB摄像头的自动驾驶汽车。您将学习实现多种自动驾驶功能，包括： * **车道保持辅助** * **巡航控制** * **T型路口导航** * **路口通行** **课程内容详述：** * **ROS2 软件包：** * 世界模型创建 * Prius OSRF Gazebo 模型编辑 * Nodes 和 Launch Files * Gazebo 中的 SDF（场景描述文件） * SDF 中的纹理和插件 * **软件部分：** * **感知管线设置：** * 使用计算机视觉技术进行车道检测 * 使用自定义卷积神经网络 (CNN) 进行交通标志分类 * 使用 Haar 级联进行交通信号灯检测 * 使用光流法进行交通标志和交通信号灯跟踪 * **基于规则的控制算法** * **课程先修要求：** * **软件基础：** Ubuntu 20.04 (LTS), ROS2 Foxy Fitzroy, Python 3.6, OpenCV 4.2, TensorFlow 2.14 * **技能基础：** * 基础 ROS2 节点通信 * 基础计算机视觉知识 * Gazebo 模型创建 * 学习动力 * **课程流程（自动驾驶开发阶段）：** 1. **硬件启动：** 使用提供的3D模型和课程链接推荐的汽车零件，快速让自动驾驶汽车在树莓派上运行。 2. **硬件接口：** 将树莓派与电机和摄像头连接，开始重要的编程工作。 3. **概论与趋势：** 理解自动驾驶的概念，以及它如何改变交通运输和环境。 4. **行业对比：** 对比特斯拉和 Waymo 等自动驾驶巨头。 5. **成果展示（模拟）：** 通过模拟环境直接展示课程成果。 * **核心功能模块：** * **车道保持辅助 (Lane Assist)** * **巡航控制 (Cruise Control)** * **T型路口导航 (Navigating T-Junction)** * **交叉口通行 (Crossing Intersection)** * 每个功能开发都包含 **检测 (Detection)** 和 **控制 (Control)** 两部分： * **检测：** 收集该功能所需信息。 * **控制：** 根据接收到的信息提出适当的响应。 * **软件要求回顾：** Ubuntu 20.4, ROS2 Foxy, Python 3.6, OpenCV 4.2, TensorFlow。 * **必备素质：** 对大型编程项目充满动力。 * **获取代码：** 建议在购买前查看课程的 GitHub 仓库，或者至少获取代码进行学习。
This Course Contains ROS2 Based self-driving car through an RGB camera, created from scratch Self Drive Features:- Lane Assist- Cruise Control- T-Junction Navigation- Crossing IntersectionsRos PackageWorld Models CreationPrius OSRF gazebo Model EditingNodes, Launch FilesSDF through GazeboTextures and Plugins in SDFSoftware Part:Perception Pipeline setupLane Detection with Computer Vision TechniquesSign Classification using (custom-built) CNNTraffic Light Detection Using Haar CascadesSign and Traffic Light Tracking using Optical FlowRule-Based Control AlgorithmsPre-Course RequirmentsSoftware BasedUbuntu 20.04 (LTS)ROS2 - Foxy FitzroyPython 3.6Opencv 4.2Tensorflow 2.14Skill BasedBasic ROS2 Nodes CommunicationBasic CV knowledgeLaunch FilesGazebo Model CreationMotivated mind:)Course Flow (Self-Driving [Development Stage])We will quickly get our car running on Raspberry Pi by utilizing 3D models ( provided in the repository) and car parts bought from links provided by instructors. After that, we will interface raspberry Pi with Motors and the camera to get started with Serious programming.Then by understanding the concept of self-drive and how it will transform our near future in the field of transportation and the environment. Then we will perform a comparison between two SD Giants (Tesla & Waymo) ;). After that, we will put forward our proposal by directly talking you inside the simulation so that you can witness course outcomes yourself.Primarily our Self Driving car will be composed of four key features. 1) Lane Assist 2) Cruise Control 3) Navigating T-Junction 4) Crossing IntersectionEach feature development will comprise of two partsa) Detection: Gathering information required for that featureb) Control: Proposing appropriate response for the information receivedSoftware Requirements Ubuntu 20.4 and ROS2 Foxy Python 3.6OpenCV 4.2TensorFlowMotivated mind for a huge programming Project- Before buying take a look into this course Github repository or message ( if you do not want to buy get the code at least and learn from it:) )
