DOC PREVIEW
Columbia CSEE 4840 - Pivoting Object Tracking System

This preview shows page 1 out of 2 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 2 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 2 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

IntroductionMain DeliverablePlan of ActionMilestonesReferencePivoting Object Tracking System[CSEE 4840 Project Proposal - March 2009]Damian AncukiewiczApplied Physics and Applied [email protected] RoyComputer Science [email protected] ShaoComputer Science [email protected] ShenElectrical Engineering [email protected] main goal of this project is to develop an objecttracking system that enable a camera, which is placed on amobile robot platform, to track a moving object. This projectcombines software (in C) and hardware (in VHDL) design,together with the knowledge of computer vision and roboticscontrol.Keywords Tracking, Object Recognition, iRobot, Wiimote,DE2, Nios II1. IntroductionObject recognition is a common application of computervision, and can be used to track the position of a movingobject with a camera. This project aims to implement thisand also to extend it by allowing the camera to pivot at will,in order to always keep the object in question in view.2. Main DeliverableThe main deliverable of this project will be a working sys-tem with a controlled (by keyboard or Wiimote controller)camera that tracks moving objects and displays the trackedimage on VGA. Since we have access to an iRobot Createrobot, we will use it as a convenient platform for pivotingthe camera horizontally. Depending on available resources,vertical pivoting might be implemented as well.The user interface will allow the user to manually panthe camera and select a target object, at which point theplatform will attempt to always keep the object in the centerof its field of view. An ambitious feature we might wantto include in our project is to use a Wiimote instead of akeyboard to choose an object for tracing. The difficulties liesin that Wiimote uses Bluetooth for its communication, whichmeans that a Bluetooth USB dongle would need to be usedand a device driver for it would need to be written.3. Plan of ActionDeveloping this project will involve four main chal-lenges: finding a way to pivot the camera, taking in videofrom the camera, recognizing a given object, and addingan interface to manually move the camera and designate anobject.Action 1The iRobot Create will be used to pivot the camera hor-izontally (yaw). The camera will be placed on top of therobot, and the Altera DE2 will command the robot to turnusing its RS-232 serial port. Vertical pivoting (pitch) will bemore difficult, as the robot cannot do that on its own. Theimplementation of this will depend on whether we are ableto find hardware that will accomplish this task.Action 2A digital camera will be used to capture video. The cam-era will output standard NTSC composite video. The DE2has an Analog Devices ADV7181 TV decoder chip whichcan convert this to a digital format. We will need to find outhow to interface the chip with the FPGA, configure it prop-erly, and access the data with the Nios II.Action 3Object recognition will be accomplished through soft-ware on the Nios II. Although several computer vision meth-ods can be utilized, the most promising seems to be scale-invariant feature transform (SIFT) in which the scales andrelative distances of features on the target image are char-acterized and then compared to those on candidate images.The advantage of this method is that it can still recognizea given object if the distance and orientation change. If itturns out that this algorithm is too intensive to be carried outin real time on the Nios II, a simpler algorithm will need tobe used.CSEE 4840 Embedded System Design, Spring 2009 1 2009/3/2Action 4An interface will need to be developed to allow the userto select a target and to see what the camera is tracking. TheDE2’s VGA and PS/2 ports can be used to connect a monitorand keyboard respectively. If we decide to use a Wiimote, wewill need to obtain a USB Bluetooth dongle, write a devicedriver for the dongle, and find out how to interface with theWiimote using Bluetooth.4. MilestonesMilestone 1(March 30)The first milestone will define the architecture and inter-face of each component. The system will be able to displayvideo from the camera. Additionally, it will be able to com-mand the iRobot to move left and right, possibly by key-board.Milestone 2 (April 13)By then, the system will have most of its features, includ-ing object recognition and controlling the iRobot Create totrack the moving object.Milestone 3 (April 29)After the second milestone, we will try our best to im-plement two more interesting, though not essential, features.The first one is moving the camera up and down and not justleft and right. Unfortunately, as of now we are not sure ifthere is such a device we can use. We will talk to Prof. PeterAllen next week and hopefully we will get any useful infor-mation from him. The second feature is to use a Wiimoteinstead of a keyboard for control. As mentioned previously,however, the work necessary to implement this is not trivial.5. Reference1. ADV7181 datasheet2. Information on the iRobot Create3. Introduction to SIFTCSEE 4840 Embedded System Design, Spring 2009 2


View Full Document

Columbia CSEE 4840 - Pivoting Object Tracking System

Documents in this Course
SPYCAM

SPYCAM

91 pages

PAC-XON

PAC-XON

105 pages

lab 1

lab 1

6 pages

memory

memory

3 pages

Structure

Structure

12 pages

Video

Video

3 pages

pacman

pacman

4 pages

Lab 1

Lab 1

6 pages

Scorched

Scorched

64 pages

lab 1

lab 1

3 pages

Video

Video

22 pages

Memory

Memory

23 pages

DVoiceR

DVoiceR

29 pages

MAZE

MAZE

56 pages

PAC XON

PAC XON

13 pages

PACXON

PACXON

13 pages

MP3 Player

MP3 Player

133 pages

Load more
Download Pivoting Object Tracking System
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Pivoting Object Tracking System and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Pivoting Object Tracking System 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?