Homework 6: Color Detection and Movement Part 1: Run the move_base_client example on the real robot. This part is ungraded. To run it, unzip the archive file in ~/catkin_ws/src and compile the workspace. In a new terminal, run “rosrun move_base_client move_base_client” to run the node. Part 2: Color Detection Write a ROS node which subscribes to the image topic /nav_kinect/rgb/image_color. As with the example from class, your node should open two windows, one to show the input image and one to show the output. Your task is to use color filtering to detect the pink hat in the image. The output image can show the segmented parts of the image which contain the hat and an addition, you must draw a circle at the x,y location where your code thinks the center of mass of the hat is. If the hat is not present in the image, then ideally your program should be able to detect that it is missing; in that case, do not draw a circle. To test your program, you can use rosbag to play a few recorded bag files that have been pre-recorded. To play a bag file, the terminal command is: rosbag play -l You must execute that command in the same working directory where the bag files are located, otherwise the path to the filename would need to be an absolute path. You can download the recorded bags from: https://drive.google.com/file/d/0B0iAtDPYl4EuS2NGRXpOOGs1bmc/view?usp=sharing After testing using the provided bag files, you must also test on the real robot. When testing on the real robot, there is no need to play any bag files as images from the robot’s camera will already be published on the topic /nav_kinect/rgb/image_color Part 3: For the last part, your task is to turn the robot in the direction of the hat. For this part, you’ll have to test on the real robot. Use your solution from Part 2 to estimate an x,y position of the hat in the image. After the hat has been detected, use the example code in move_base_client.cpp to turn the robot towards the hat. The goal message should contain 0s for both x and y and only modify the desired angle. WHAT TO TURN IN: A single ROS package called _homework6 where you substitute your EID in place of . Include a README file with instructions on how to run Part 2 and Part 3 (these should be two separate nodes). For Part 2, include 5 different screenshots which show the input and output windows with a circle drawn around the hat in the output windows. I recommend that in the output window, you set all pixels not belonging to the hat to 0 RGB so that the only pixels that show anything are those matching the color of the hat. For Part 3, you will have to demonstrate your solution a mentor or the TA to get full credit. When you demonstrate this, upon logging in the robot’s computer, download your package from the canvas website (the exact same thing you submitted). It’s ok to do the demonstration after the due date as long as the mentor clearly sees that you’re downloading the package that you turned in by the due date. Good luck!