
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Messages  yashardel
1
« on: July 12, 2009, 02:38:41 PM »
Hi !
I am writing a report about a project I have recently done on Wii Remotebased head tracking. I need some information about the bandpass filter that is used in Wiimote's camera, information like the peak frequency, the bandwidth of the filter etc.
Are there some documentations on this or any ideas about filter approximate properties?
Thanks.
Best Regards, Yashar
3
« on: July 11, 2009, 06:46:39 PM »
Hi ! Thanks for your response. From experimental point of view that is right. The effective FOV might be level less that but what I meant is to find a relationship between the FOVx and FOVy theoretically. It means , if these two assumption below hold true FOVy can NOT exceed the maximum doublesided FOV of 33.75 degrees. The assumptions are :  px=py (same pixel size in x and y direction) !!
 Resolution 1024*768
4
« on: July 11, 2009, 06:21:44 PM »
Hi everybody.
I have an idea how to estmate the camera's vertical field of view.
It is said that the Wii IR camera has an ideal Field Of View (FOV) horizontally (pi/4)=45 degrees .
Assuming that the camera has 1:1 ratio of the pixel size meaning px=py then and knowledge of the resolution 1024*768 pixels gives us :
px . . . . . py . . . . .
. . . . .
. . . . .
px=(FOVx)/1024 = (pi/4)/1024 [rad/pixel] > this is the length of one unit of pixel in x direction
py=(FOVy)/768
px=py > FOVy/768=(pi/4)/1024 > FOVy=(3/4)*(pi/4) [rad] =33.75 degrees
Does anybody have any idea about this method!?
BR, Yashar
5
« on: February 25, 2009, 10:26:14 AM »
Hi! I am wroking for wiiremote tracking project and now we are standing in a point which we need the pixelsize and focal length in Milimeter; The point is we are able to estimate these values in respect to each other as :
assume : pixel size (1.0,1.0) focal length h=12801300 pixel size
But as you see they are related to each other. We need to know each os these two parameters soly interm of pixel? Please help us if you can!
Thanks ...
6
« on: February 25, 2009, 04:41:48 AM »
Yeah I see what you mean! Well the problem is almost solved. The things is when you are evaluating the angles in theroy in order to compare them with result of your measuremnt, the accuray needs to be so high for low distances(or low angles) and even a slight tolerance or range 0.2 mm can have big affect(e.g 10 times) on the final evaluated distances(r1,r2,r3).
7
« on: February 24, 2009, 08:35:59 AM »
Hi! I am working on wiiremote based headtracking . There have been a small problem in the measurment of angles which is bothering me! To simlpify my words, each time when my LED alighemnts is in a way that two adjacent LEDs are so closed to each other (and so the angle between the vectors from the focal point to the points is SMALL) the error ofthe measurement incredibly increases and the angles are evaluated wrongly and so I cannot trust my mesurement anymore. Has anybody als ofaced before such a problem?
8
« on: February 07, 2009, 05:46:44 PM »
Hi!
I have been contemplating using 1 wiimote for headtracking applications. But seems that with 1 wiimote , getting full 6DOF info is rather difficult due to the big mathermatics behind. Does anybody know if using 2 wiimote have any advantage over 1 wiimote ? Do we get any further information combing two wiimotes or no it's just concidered as two seperate measurement from different angles which are combined together and the result is a more robust and accurate estimate?!!
Thanks for assistance!
9
« on: January 29, 2009, 10:57:46 AM »
Hi! I am looking for some LED array like the one John lee is using in his VRDesktop headttracking. Do you know where it is possible to order one ? (No doubts the LEDs should be IRleds that are used for wiimote.)
Thanks!
10
« on: January 29, 2009, 08:01:59 AM »
Ok!thanks ,yeah it is really small. Perhaps that explains why incorporating two wiiuse gives more robust estimation!
11
« on: January 29, 2009, 04:19:51 AM »
Hi! I was just wondering when in the documentation it says the camera's field of view is 45 degree, is it half of a sysmetirc angle meaning 90 dergree in total or not in totat it is 45 degree? If it's 45 degree in total then it should be such a small range! What do you think!?
12
« on: January 27, 2009, 07:03:44 AM »
Hi! There is a point in John lee's code that I can't understand well.
radiansPerPixel is defined as (pi/4)/(1024) for both verstical and horizontal case. We know that the camera's resolution is 1024*768 so shouldn't in vertical case we use (pi/4)/(768) instead!!!?
13
« on: January 27, 2009, 06:38:25 AM »
By the way, when the camera is above the acreen 0.5f is added not subtracted!?You know why!?
if(cameraIsAboveScreen) headY = .5f+(float)(movementScaling * Math.Sin(relativeVerticalAngle + cameraVerticaleAngle) *headDist);
14
« on: January 27, 2009, 06:28:34 AM »
Hi!
To Wilco: Thanks man! That explanation really helped!
But I have querstion too. If you have the sketch of Horizontal and Vertical Position in front of you, would see that headDist is different in two positions but why they are used as being similar! I guessit is a mistake unless I am ignoring something! The relation is as below:
headDist_Y^2=headDist_X^2+DheadY^2 or similarly
tan(teta_headY)=DheadY/headDist_Y where headDist_Y is not what was previously calculated as headDist_X;
Can you help me?!
15
« on: January 27, 2009, 05:54:59 AM »
Hi ! There are few important question in John Lee's code that I can't understand ! Can you help me with them ?!...
Code: relativeVerticalAngle = (avgY  384) * radiansPerPixel;//relative angle to camera axis
if(cameraIsAboveScreen) headY = .5f+(float)(movementScaling * Math.Sin(relativeVerticalAngle + cameraVerticaleAngle) *headDist); else headY = .5f + (float)(movementScaling * Math.Sin(relativeVerticalAngle + cameraVerticaleAngle) * headDist);
 The headDist calculated in Vertical position differs with that in Horizontal position. John is using the same headDist in both cases? Isn't he doing that wrong...?! headDist_Y^2=DheadY^2+headDist_X^2 where headDist_X is the head distance calcualted in the Horizontal poistion.Why is he ignoring this simple fact!!!!?
 What is the effect adding +/ 0.5 f to the formula?
 Why do we need to write (AvgY384) and NOT only Avg Y? is the subtraction necessary?
Hope some body can assist me alittle!


