This repository contains the Go Chase It project from the Udacity Robotics Software Engineer Nanodegree.

In this project, a differential-drive robot is spawned in Gazebo with a camera and lidar. A C++ perception node analyzes camera images to detect a white ball and commands the robot to chase it through a ROS service.

• Added perimeter boundary walls in the active world to keep the simulation fully enclosed.
• Lowered white-ball detection threshold from 255 to 200 for more reliable detection under varied lighting.
• Improved process_image control logic to use centroid-based steering and smoother velocity commands while approaching the ball.

The workspace includes these ROS packages under catkin_ws/src:

• my_robot
• ball_chaser

The my_robot package contains the robot model and simulation world setup.

Completed work:

• Built a differential-drive robot using URDF/Xacro.
• Added two sensors: camera and lidar.
• Added Gazebo plugins for:
  • Differential drive control
  • Lidar sensor
  • Camera sensor
• Added launch files to load the robot description and start Gazebo world simulation.

The ball_chaser package contains service definitions and C++ nodes for chasing behavior.

Completed work:

• Created DriveToTarget.srv service.
• Implemented drive_bot node:
  • Provides /ball_chaser/command_robot service.
  • Receives requested linear_x and angular_z values.
  • Publishes velocity commands to /cmd_vel.
  • Returns feedback with commanded values.
• Implemented process_image node:
  • Subscribes to /camera/rgb/image_raw.
  • Detects white pixels corresponding to the white ball.
  • Determines whether the ball is left, center, or right in the image.
  • Calls the service client to steer robot toward the ball.
• Added ball_chaser.launch to run both nodes:
  • drive_bot
  • process_image

go-chase-it/
  catkin_ws/
    src/
      my_robot/
      ball_chaser/

From the project root:

cd catkin_ws
catkin_make

Source the workspace:

source devel/setup.bash

Open terminal 1 and launch the world and robot:

cd catkin_ws
source devel/setup.bash
roslaunch my_robot world.launch

Open terminal 2 and start ball-chasing nodes:

cd catkin_ws
source devel/setup.bash
roslaunch ball_chaser ball_chaser.launch

• The robot camera detects the white ball.
• The process_image node detects bright ball pixels (threshold 200) and computes the ball centroid in the image.
• The node requests /ball_chaser/command_robot service.
• The drive_bot node publishes smooth linear and angular commands to /cmd_vel.
• The robot steers proportionally toward the ball, slows while turning, and stops if the ball is not visible.