Nano入门教程软件篇-安装AI学习库ROS接口ros_deep_learning

Nano入门教程软件篇-安装AI学习库ROS接口ros_deep_learning

说明：

• 介绍如何在nano上安装AI学习库ROS接口ros_deep_learning

步骤：

• 首先安装ros melodic， 参考教程

• 安装jetson-inference ，参考教程

• 安装ros_deep_learning

$ cd ~/catkin_ws/src
$ git clone https://github.com/dusty-nv/ros_deep_learning
$ cd ../
$ catkin_make

imageNet测试：

• 新窗口，启动roscore

roscore

• 新窗口，发布图片话题

$ rosrun image_publisher image_publisher __name:=image_publisher ~/jetson-inference/data/images/orange_0.jpg

• 新窗口，启动识别

$ source ~/catkin_ws/devel/setup.bash
$ rosrun ros_deep_learning imagenet /imagenet/image_in:=/image_publisher/image_raw _model_name:=googlenet

• 新窗口，结果显示

$ rostopic echo /imagenet/classification

detectNet测试：

• 关闭上面所有窗口

• 新窗口，启动roscore

roscore

• 新窗口，发布图片话题

$ rosrun image_publisher image_publisher __name:=image_publisher ~/jetson-inference/data/images/peds-004.jpg 

• 新窗口，启动识别

$ rosrun ros_deep_learning detectnet /detectnet/image_in:=/image_publisher/image_raw _model_name:=pednet

• 查看结果：

$ rostopic echo /detectnet/detections

参考：

• https://github.com/dusty-nv/ros_deep_learning