Download Key Task Areas

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Automated airport weather station wikipedia , lookup

Transcript
Mobile Sensing – Plans and Tasks
Key Task Areas:





Leaf Inspection – Vision (Alex Han)
Plant Detection – 3D Vision
World Mapping – Plant Locations
Path Planning – Navigate Around the Plants
Robot Control – Path Execution
Progress(0-5)
:
0.1
0
0
0
0
Task Plan:
1.
2.
3.
4.
5.
…
Team Selection
Computer Setup
Full Computer Environment Setup, and Software Examples.
Perception: Collect Images of Single Leaves, Experiment with Detection Methods.
Controls: Use a ROS Simulation with Simulated Sensor Data, use high/low level control.
Name1:
Get ROS Turtle Bot Laser Sim Running. (Task for this week)
Got it working, but don’t understand the code. (End of week result)
Chenming Du(Jason):
Collect Leaf Images, Start Analyzing in OpenCV.
Had camera problems, got images, no analysis yet.
Niraj Javia: Work with Gazebo and get used to the environment
Had trouble finding it and kept getting simulation error
Alex Han: Analyze sample leaf image with OpenCV to detect leaf’s edge and abnormality
Got some results with several methods
6.
Alex Han: Goal: Understand ConvexHull method and go in-depth with it for leaf analysis. Look for
other methods that might fit better for the purpose
Chenming Du Work with Gazebo and get used to the environment
Niraj Javia Tried to get stage and Gazebo working but it did not work
7. Perception: Detect defects on single leaves and multiple/overlapping leaves.
Controls: Get ROS running on the turtle bot, test controls, integrate sensors/camera.
Niraj Javia: Gazebo Gazebo Gazebo Gazebo Gazebo Gazebo Gazebo Gazebo
Trouble running a ROBOT in this silly thing! Blahhhh
Chenming Du :go through the flow chart and Gazebo
8. Chenming Du :running a Turtlebots in a circle in Gazebo
Niraj: Same as Chenming
9. Perception: Use Sensor data to detect rows of plants, filter/reduce noise, create map.
Controls: Create reactive safe behaviors using the sensor data, robot localization.
10. Chenming Du : pin the two laptops together, one is PC, one is laptop on Turtlebots. Use Joystick
Controller to control the robots.
Niraj: Assembled turtlebot and connected kinect and pinged both the turtlebot laptop and my
laptop
11. Perception: Detect defects on overlapping leaves, at different scales, local/global planning.
Niraj: Ran the follower demo successfully! Had trouble gathering images from kinect and controlling
turtle bot with keyboard
Chenming Du: Use the keyboard to control the turtlebot through ROS topic
12.
13.
14. Demo: TurtleBot Driving down a row of plants detecting a plant defect.