Sponsors of Wiimoteproject.com
IR Pens for Wiimote Whiteboard
Wiimote Project » Forum » General Wiimote and Technology » Project Ideas » get the 3D coordinates to localize target for medical imaging using wiimote
Pages: [1]
« previous next »
  Print  
Author Topic: get the 3D coordinates to localize target for medical imaging using wiimote  (Read 2528 times)
0 Members and 1 Guest are viewing this topic.
Irfan syed
*
Karma: +0/-0
Posts: 1
Offline Offline
View Profile Email
« on: April 19, 2008, 05:28:24 PM »
Hi all
I am trying to use wiimote to 3D coordinates of sources and detectors that emit infrared from patient's head for medical imaging.
I don't know much about C# and all of my programing is Matlab. I was going thru the code of Johnny lee's 3D head tracking, but i didn't get how he is getting the follwoing
1) z coordinate
2) what is the reference origin
I also have few other questions, but let me begin by finding the algorithm
first
thanks for help in advance guys
Irfan
Logged
ZaPPZion
Guru
*
Karma: +11/-1
Posts: 137
Offline Offline
View Profile
« Reply #1 on: April 20, 2008, 09:40:36 AM »
It's quite hard to get 3D positions of an IR LED with just 1 wiimote. It's possible to get the size of the IR blob that's being tracked, but that doesn't say everything about the z-axis (if you mean the axis that's pointing away from you, so distance to the screen). If you tilt the IR LED, the intensity will decrease as well, so it's not accurate at all. I think it would be better to use 2 wiimotes. 1 for the x/y picture, and 1 for the y/z picture for example. That's already been done and shouldn't be too hard to implement (i think)
Logged
yashardel
*
Karma: +0/-0
Posts: 37
Offline Offline
View Profile
« Reply #2 on: January 21, 2009, 04:10:41 AM »
 Hi!
I have the same problem how to find the z coordinate using a 2D IR camera. I am using the wiiuse code written in c to do my headtracking. As you said, there is a varaible in the IR data structure named IR_size (0-15) which I guess is responsible for calculating the z-coordinate(distance from the screen) but it's not being used. Now my question is , how did John Lee accompolish the heatracking solution using only 1 wiimote? The VR-Desktop he has made works beautifully and he used 1 signle wiimote with some IR-LED...

What do you think people...? Can you share your ideas?

Thanks...
Logged
boonjin
*
Karma: +23/-0
Posts: 479
Offline Offline
Smoothboard Developer
View Profile WWW
« Reply #3 on: January 21, 2009, 07:25:21 AM »
Quote from: yashardel on January 21, 2009, 04:10:41 AM
Hi!
I have the same problem how to find the z coordinate using a 2D IR camera. I am using the wiiuse code written in c to do my headtracking. As you said, there is a varaible in the IR data structure named IR_size (0-15) which I guess is responsible for calculating the z-coordinate(distance from the screen) but it's not being used. Now my question is , how did John Lee accompolish the heatracking solution using only 1 wiimote? The VR-Desktop he has made works beautifully and he used 1 signle wiimote with some IR-LED...

What do you think people...? Can you share your ideas?

Thanks...

Hi,

Johnny used two infrared LEDs to be placed at the googles/specs. With the detected distance between the IR points, you can estimate the distance between the user and the screen.

Regards,
Boon Jin
Logged
Smoothboard.net
Wii Whiteboard/Wiimote Whiteboard with annotation features, autoconnect to multiple Wiimotes and more!

Smoothboard/Wii Whiteboard Wiki
One stop guide to all your Wiimote Whiteboard needs

BoonJin.com
insertnick
*
Karma: +0/-0
Posts: 35
Offline Offline
View Profile
« Reply #4 on: January 21, 2009, 07:34:07 AM »
See also free-track project (3 leds)
(they say wiimote support coming soon).
Probably with three leds you can track well also tilt and rotation movements, while with two you can only "scroll" your head left and right (I suppose).
« Last Edit: January 21, 2009, 07:36:18 AM by insertnick » Logged
yashardel
*
Karma: +0/-0
Posts: 37
Offline Offline
View Profile
« Reply #5 on: January 21, 2009, 09:16:17 AM »
 Hi !
Thanks for the information.

Regarding Boon Jin comment, the question is how the 3D localization is achieved from 2D info? I don't think distance between LEDs (which we know by default) and x-y location info suffices for this. I believe he is using a more complex method namely Perspective transformtion to calculate the depth information. (i.e the distance) But I am not familair with this method? Any body can assist me a little?

Logged
Marco
*
Karma: +0/-0
Posts: 3
Offline Offline
View Profile Email
« Reply #6 on: January 22, 2009, 04:56:09 AM »
Hi,
I've been looking at is as well, this is what I came up with.

3D tracking with 1 Wiimote
Yes it is possible but you need 3 or better 4 IR points in a known configuration. See this rather hefty math article to get an idea: www.infoweb.drake.edu/mathcs/rieck/wii.pdf .

3D tracking with 2 Wiimotes
This is easier, at least the math is  Cheesy.
You place 2 Wiimotes on the same line (base of a triangle is will form) at a known angle from this line. Remember the angle that the Wiimotes can see if about 40 degrees.
Then you use the X coordinate of both Wiimotes to determine the angle from the baseline.
Then you use triangulation to calculate the location of the point in the XY plane (the plane parallel to the floor). You can use the Y to do the Z.

Marco
Logged
Pages: [1]
  Print  
« previous next »
 
Jump to:  

TinyPortal v.1.0.6 beta 2 © Bloc
Clicky Web Analytics