Setup

For this lab, get the robot to do something interesting using vision. You will need to use one of the Magellan's for this lab as they are the only ones with onboard cameras. Alternatively, you could try using the overhead camera installed in the Mural room with the Nomad running around within view of the camera.

Checkout svm from /usr/local/cvs on the engin server. The two binaries you will want to be using are svm and dispipc. The IPC central server must be running before you start up svm. When you start svm, it looks for two files in the current directory, camera0.conf and camera1.conf. Those are located in the modules subdirectory of the svm code tree. The configuration files determine which cameras are active and whether the PTZ control is activated. The source file dispipc.c has examples of how to subscribe to svm data messages that indicate when something is identified by the vision model (e.g. a face or a pink badge). It also contains examples of how to turn operators on and off using IPC messages.

You will want your arbiter or control module to subscribe to svm data messages and turn on the desired operators. Then you can use a state machine to manage the behavior of the robot when it receives visual cues.

Task:

Pick one of the following tasks and integrate it with your robot system. For this lab you will have to work exclusively on the robots, as the simulator cannot produce simulated video. It is important, therefore, that you have the robot spitting out lots of debugging information as it moves so that you can figure out what it is doing. A log file is helpful, as is the use of spoken information as the robot is moving around. We can put speakers on the robots for that purpose, and you can use festival to generate synthetic speech.

  1. Use the face detector and/or skin detector to find people and greet them appropriately while wandering around Hicks. If you can remember people using histograms, so much the better. Motion is also useful for this task.
  2. Have the robot follow a marker, such as the blue wand in my office. If it can't find the marker, it should wander around looking for it.
  3. Have the robot find signs in Hicks and read them to localize itself. You will need to generate the signs. Have it undertake a simple task like delivering mail from the office to other offices on the first floor, although just wandering around is fine..
  4. Use p-similar patterns pasted around Hicks to let the robot know where it is. An office delivery task is appropriate here as well, but just wandering is fine.
  5. Implement a line-following algorithm, and use it to have the robot track a line on the ground to get between various locations in Hicks. This will require creating a new operator in SVM (Zach P. may have one built already).
  6. Have the robot track and "kick" a soccer ball. In a desire direction.
  7. Do visual localization using colored markers or features of the environment.
  8. Something else interesting.

Writeup

All of the writeups will be submitted as web pages. When you are ready to submit your lab, use the on-line submission form. Code should be kept on an isolated link or be password protected and linked to the lab writeup. The writeups should be well-written, complete, concise, and organized. Images and pictures should be integrated with the text of the writeup in an effective manner and should not dominate the page. The best writeups make good tutorials on the topic, explaining the algorithms and providing examples of them working. Ten percent of the lab grade will be based on the quality of the writeup.

All writeups should contain at least the following elements.