more stable realase of the sdk

[ahmad2...@yahoo.com]

first of all thank you for the great sdk that i have been working on
for six month almost
now we are working on phyiscal therapy game for childs and after going
far in this project and got a good achivment in is filed we are going
to Imagine cup .

but now we are at a level that your sdk can't give us good matching of
points and the and it has a lot of disstortion

so i have to ask before we change to ms kienect sdk are you going to
relase a more stable relase of this sdk with more features and if yes
i am willing to share our achivment with you



thank very much

[Dan Amerson]

Ahmad,

We're very glad that you've had success with ICM. Your project sounds
interesting, and we hope you do well in the Imagine Cup with or without ICM.

We have not announced plans to release a new version of ICM at this time.

If you can elaborate on what you mean by matching of points and distortion,
perhaps we can help you improve your results with the current version. Are
you seeing that your motions don't map well to the actor? Are you concerned
about latency? What is the use case where ICM is failing for you?

Thanks,

dba

[ahmad2...@yahoo.com]

hI I FORGET TO SEE HAPPY NEW YEAR FOR YOU AND YOUR TEAM
SECOND THING MY PROBLEM HAPPENS IN EVENT GENERATOR THE MATCHING HAS A
LOT OF PROBLEM CAN'T BE USE IN PHYISCAL THERAPY GAME
AND ANY DISSTORTION CAN EFFECT IT LIGHTS OR NOISE OF ANY KIND.

AND ONE QUESTION DOSE YOUR SDK CAN BENEFIT FORM THE UPDATES OF OPEN NI
MODULES OR NOT


THANKS

On Jan 2, 6:54 pm, "Dan Amerson" <amer...@activate3d.com> wrote:
> Ahmad,
>
> We're very glad that you've had success with ICM. Your project sounds
> interesting, and we hope you do well in the Imagine Cup with or without ICM.
>
> We have not announced plans to release a new version of ICM at this time.
>
> If you can elaborate on what you mean by matching of points and distortion,
> perhaps we can help you improve your results with the current version. Are
> you seeing that your motions don't map well to the actor? Are you concerned
> about latency? What is the use case where ICM is failing for you?
>
> Thanks,
>
> dba
>
>
>
>
>
>
>
> -----Original Message-----
> From: activat...@googlegroups.com
>

> [mailto:activat...@googlegroups.com] On Behalf Of ahmad2382...@yahoo.com

[Dan Amerson]

ICM should be compatible with newer versions of OpenNI and NITE since the
interfaces to the DLLs do not change. I have ICM 1.1.1 installed, and it
works correctly with OpenNI 1.3.3.6 and NITE 1.4.2.4. These are the latest
stable versions from www.openni.org.

Matching will vary from gesture to gesture based on the type of motion.
Cyclic motions that involve end-effector motion, hands and feet, should
match best.

Unfortunately, if you're getting interference from light (presumably
sunlight), there's little we can do at the software level. Sunlight
interferes with the depth measurements of the sensor which in turn corrupts
the depth map and skeleton data. The best we can do is try to filter the
joint positions to remove some jitter and detect truly incorrect information
from NITE.

[ahmad2...@yahoo.com]

yeah i know it work with the new open ni but dose it take with the new
features form the new onse like better detection and so on

where to modify the C# for depth image and kinect thing



thanks :)

On Jan 2, 8:41 pm, "Dan Amerson" <amer...@activate3d.com> wrote:
> ICM should be compatible with newer versions of OpenNI and NITE since the
> interfaces to the DLLs do not change. I have ICM 1.1.1 installed, and it
> works correctly with OpenNI 1.3.3.6 and NITE 1.4.2.4. These are the latest
> stable versions fromwww.openni.org.
>
> Matching will vary from gesture to gesture based on the type of motion.
> Cyclic motions that involve end-effector motion, hands and feet, should
> match best.
>
> Unfortunately, if you're getting interference from light (presumably
> sunlight), there's little we can do at the software level. Sunlight
> interferes with the depth measurements of the sensor which in turn corrupts
> the depth map and skeleton data. The best we can do is try to filter the
> joint positions to remove some jitter and detect truly incorrect information
> from NITE.
>
> dba
>
>
>
>
>
>
>
> -----Original Message-----
> From: activat...@googlegroups.com
>

> [mailto:activat...@googlegroups.com] On Behalf Of ahmad2382...@yahoo.com

[Dan]

We use the skeleton data from NITE, so any improvements in tracking
that come with new versions of NITE will translate directly to
improvements in ICM.

I'm not sure what depth image and Kinect thing you are inquiring about
given the context of my previous message. Can you clarify?

dba

On Jan 2, 1:46 pm, "ahmad2382...@yahoo.com" <ahmad2382...@yahoo.com>
wrote: