AIM Develop a camera guided robot Has to be able to detect red circular objects and the object position in each frame Needs to translate image coordinates to real world coordinates for robot ID: 809708
Download The PPT/PDF document "Camera Aided Robot Progress Report" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Camera Aided Robot
Progress Report
Slide2AIM?
Develop a camera guided robot.
Has to be able to detect red circular objects and the object position in each frame.
Needs to translate image co-ordinates to real world co-ordinates for robot.
R
obot uses the real world information to locate object.
Draws a dot on centre of object.
Slide3Block Diagram
4460 Core
A
rduino
Dagu
4 channel motor driver
Pandaboard
Dagu
Rover 5
USB
Motor controls
encoders
Motor 1
Motor 2
Motor 3
M1OUT
M2OUT
M3OUT
Camera
USB
Slide4System Flow
Start
Initialize camera and robot
Start looking for objects
R
ight
Detect position of object
object found?
L
eft
Middle
object found?
Turn robot left
Stop robot
Turn robot right
Drive forward
Y
es
No
No
No
Y
es
Y
es
Y
es
Slide5Development Environment
Hardware Development Board
The development board chosen is the Pandaboard ES.
1.2GHz Dual core Arm
9 processor,
SGX540 graphics core which
supporting all major API's.
1GB RAM, Adequate memory.
Slide6Development
Enviroment
Logitech 330 HD webcam being used for camera.
Operating System
The OS chosen for this
project is Ubuntu OMAP4.
TI kernel for the Pandaboard ES.
Support for OpenCV .
Driver support for camera’s (v4l) and arduino.
Slide7Development
Environment
Robot
To control the robot I have
decided
to use an
A
rduino
for
the sole control of the robot.Dual motor driver from Pololu with two MC332926 chips has been chosen as the motor driver.
Dagu Rover 5 is used for the robot chassis.
Slide8Progress To Date
Pandaboard And Ubuntu Issues
Difficult integration of Ubuntu OS and the Pandaboard.
TI kernel
ondemand
issue with clock solved.
Boot.script issue with RAM solved.
Slide9Progress To Date
Application
Finds red circular objects in an image.
Elliptic Morphological
kernel
used to
manipulate image to find circular objects.CvBlobs library used to extract circular objects which are then stored in
arrays.Kalman filter used for tracking the object based on centre point of object. noise and other external factors need to be taken into account.
Slide10Progress To Date
Yellow x is the predicted centre of the object.
Blue square is the bounding box drawn by CvBlobs library.
The Memory used is small (Just 15 – 25 MB).
Slide11Progress To Date
Robot
& Arduino
Robot is not working.
Simple code
developed for Arduino to control the robot.PWM generated depending on character received over USB serial.
Slide12Future Plans
Once a new motor driver is gotten the
entire robot
will be set
up (excluding arm).
Possible change in motor driver.
Dagu
provide a driver with 4 channels.
Simple starting application to calibrate
the robot.Integrate current Application on Pandaboard with Arduino and chassis and check calibration.
Slide13Future Plans
Get a robotic
arm and
integrate into application.
(
Dagu
- 2DOF Arm with
Gripper and Servos)
Develop an algorithm to convert 2D image co-ordinates to real world co-ordinates.
Translate the real world co-ordinates into robotic arm commands and chassis movement.
Slide14