Book Image

Learning ROS for Robotics Programming

By : Aaron Martinez, Enrique Fernández
Book Image

Learning ROS for Robotics Programming

By: Aaron Martinez, Enrique Fernández

Overview of this book

<p>Both the amateur and the professional roboticist who has ever tried their hand at robotics programming will have faced with the cumbersome task of starting from scratch, usually reinventing the wheel. ROS comes with a great number of already working functionalities, and this book takes you from the first steps to the most elaborate designs possible within this software framework.</p> <p>"Learning ROS for Robotics Programming" is full of practical examples that will help you to understand the framework from the very beginning. Build your own robot applications in a simulated environment and share your knowledge with the large community supporting ROS.</p> <p>"Learning ROS for Robotics Programming" starts with the basic concepts and usage of ROS in a very straightforward and practical manner. It is a painless introduction to the fascinating world of robotics, covering sensor integration, modeling, simulation, computer vision, and navigation algorithms, among other topics.</p> <p>After the first two chapters, concepts like topics, messages, and nodes will become daily bread. Make your robot see with HD cameras, or navigate avoiding obstacles with range sensors. Furthermore, thanks to the contributions of the vast ROS community, your robot will be able to navigate autonomously, and even recognize and interact with you, in a matter of minutes.</p> <p>"Learning ROS for Robotics Programming" will give you all the background you need to know in order to start in the fascinating world of robotics and program your own robot. Simply, you put the limit!</p>
Table of Contents (16 chapters)
Learning ROS for Robotics Programming
Credits
About the Authors
About the Reviewers
www.PacktPub.com
Preface
Index

Chapter 6. Computer Vision

In ROS, the support for computer vision is provided by means of camera drivers, the integration of OpenCV libraries, tools to set the frame transform (tf) of the camera optical frame with respect to our robot, and a good number of third-party tools, which comprise algorithms for visual odometry, augmented reality, object detection, and perception, among others.

The first capability that ROS offers us when working with vision is the ability to manage FireWire (IEEE1394a or IEEE1394b) cameras. Indeed, a package in the main ROS framework contains the drivers for such cameras. In the case of USB or gigabit Ethernet cameras, we must install third-party drivers or use our own. In this chapter, we will list USB drivers that come as ROS packages, and we will also provide a driver source code implemented with the OpenCV video capture API. USB cameras are pretty inexpensive and easy to find on the market, so they follow the goal of this book: learn by practice.

This chapter...