Getting Started with Augmented Reality / FLARToolKit
Optimize and Customize the Settings
Camera Settings
The camera in flash is a processor hog so you really want to limit the amount of data being run through the camera. The default settings in your project for the camera are 640 x 480 video, with 0.5 sample rate and 30 fps.
The first thing to consider is the camera’s frame rate. Do you really need 30 fps from the camera? This would be great if it didn’t cause your scene to run at 15 fps. So why have this set so high unless you are actually getting 30fps playback. I recommend playing with this setting a little bit to find a healthy compromise between the camera performance and overall scene performance.
Next are the dimensions. One option you have is to shrink all of the video in the scene to something smaller, but is that really a great thing? I mean, half the fun with Augmented Reality is that you have a big video and you are interacting with it. So just find the smallest possible size that meets your needs.
Finally, you have the sample ratio for the camera. This is a value between 0 and 1, where 0 is terrible and 1 is amazing. Tinkering with this value, I have found that around the default of 0.5 is a good balance of quality and speed so I usually leave this value alone but 0.4 or 0.45 may help out if you are REALLY desperate for speed.
Update Frequency
If you take a look in the View class, there is an updateScene function that runs every frame. One thing that I like to do, is as follows: Once I have dropped the camera frame rate I also create a timer in the view class that runs at the same interval that the camera updates at. This means that the matrix isn’t being translated at every frame.
This may result in more choppy performance, but it may also make things a little bit smoother if you are finding that you app is really chugging along. You could also play with an easing equation on you enter frame function to compensate if this creates a choppy feel to the application.
thanks,
marc











February 7th, 2009 at 2:07 am
I bet many of PV3D developers are waiting for something like that. It looks very interesting, it’s a nice work even if its on an early state, i wonder if someday it will turn into a desktop app (maybe it could be added as an option).
April 29th, 2009 at 6:04 pm
Hi Marc
This tutorial its so awesome, its really answers me a lot of doubts about FLAR and how to handle it
I got a question.
How can I detect the angle of the container where all the items were placed, i mean, if I rotate the pattern over any angle in diferent axis(x,y,z), how can i get the currents angle(x,y,z) were the container is rotates
Best regards,
David Andres Niño
May 5th, 2009 at 10:37 pm
A error!!, please you can’t send project source files?
Spanish
Me ocurren errores cuando quiero ejecutar los archivos, me dice que las clases tienen errores, me podrias enviar el proyecto completo?
Gracias
May 6th, 2009 at 9:03 am
Hello !!
Thanks for great explainations on a difficult subject
(at least for me)
I followed all your step but I still get an error :
Error #1063: Argument count mismatch on HelloFlarView$iinit(). 3 expected, 0 found.
(translated from french sorry)
Any idea how to fix that ?
Thanks a lot !!!
May 13th, 2009 at 8:34 pm
It’s great, but when i rotate my face, it can’t tracking
July 16th, 2009 at 7:37 pm
I have two major problems following this tutorial:
1. if i chose flash 10 nothing will compile with the error Error: unknown configuration variable ‘target-player’
10.0.0
2. if i chose flash 9 i get the error FLARManagerTest_PV3D.as(44): col: 24 Error: Syntaxfehler: identifier vor lessthan erforderlich.
var patterns:Vector. = new Vector.();
^
Build halted with errors (mxmlc).
Done (1)
i guess it’s because i use newer versions, any idea how to fix this would be appreciated
August 14th, 2009 at 6:21 pm
Thanks for the tutorial. So far, so good.
August 19th, 2009 at 3:06 am
Hey. I believe that all government is evil, and that trying to improve it is largely a waste of time. Help me! Please help find sites for: Deep conditioners recipe. I found only this - best central air conditioners. Applications the sag fighter is a specific type of power conditioner used for protection against deep voltage sags deep voltage drops. In recent years, I purchased a “prize” yellow labrador puppy and began searching the market for a shampoo and conditioner that would serve that vivid memory. Thanks for the help :rolleyes:, Giva from Hungary.
August 29th, 2009 at 4:33 am
Excuse me. Computer Science is no more about computers than astronomy is about telescopes. Help me! Could you help me find sites on the: Stock to pick. I found only this - louis navellier stock picks. Barron’s agreed the shape that madoff’s tricks were most even such to basic making his player’s ivory coefficients, future stock picks. Best regards :mad:, Kachine from Cambodia.
October 19th, 2009 at 5:54 am
hello, i am tommy ,i am a student,i want to create some flash game use the face detection, but i dont know how can do it.i try serach it information in internet but i cant search the particular description to tell the face detection by flash.
could you tell me have any website have particular to tell the flash face detection ,thank you very much.
January 11th, 2010 at 7:08 pm
Would you be so kind as to share this code?
January 15th, 2010 at 1:12 pm
if you want to check out what you can do with this library I leave my projects page here http://facepong.blogspot.com/
January 24th, 2010 at 7:03 am
Hi there,
I am trying to follow your example but i get a black window (no camera input). Any idea how to fix this??
btw: for those having the lessthan error. Make sure you target Flash player 10 and also have the latest flex sdk
March 2nd, 2010 at 1:18 am
I have an error.
C:\Users\Carlos\HelloFlarToolkit\src\views\HelloFlarView.as(103): col: 108 Error: Syntax error.
var materialsList:MaterialsList = new MaterialsList({all: new FlatShadeMaterial(pointLight3D, 0xffffff, 0×000000)});
Anyone can help me?
March 2nd, 2010 at 10:10 pm
Alex, I have the same problem. I only see a black window. Did you fix the problem?
March 2nd, 2010 at 11:11 pm
looks like you are using a special character in the colour definition (0×000000), that isn’t an x that you are using. that could be the reason for that error