Touchlib updates
I have updated the touchlib bins to fix a few things. Pressing 'b' now should clear out any existing blobs which helps for those times when touchlib gets overloaded and blobs get stuck on screen (in the future hopefully I'll be able to prevent this from happenning to begin with..). It is also using the openCV find contours function to locate blobs now though I haven't done a side by side comparision to see if it's any faster..
I also loaded the flash source into the google code repository in the flashSrc directory. This will be the preferred way to get the flash source from now on.
Bins for the Flash apps are now available here. I made some big updates to the photo app - pictures now slide around with some velocity after you let go and I fixed a bug with the two finger dragging behavior (when you let go with one finger it would snap to the wrong location before).
In other news, my employer (Archer Group) is having me build a table for them, so I'll be able to use a decent camera this time (we have ordered the FireFly MV from pt-grey research) and produce something that is a finished table, rather than a prototype. We are going for a table form factor.
The search for the perfect 'compliant surface' continues (someday we'll have to come up with a better name for it). Rockit (on the nui group forums) has reported some success with silicone rubber and mylar drafting sheets. I have a big piece of transparent silicone on order - I will let you know how well it works when I receive it. I'll be trying some of the rosco screen on the silicon rubber.
I also loaded the flash source into the google code repository in the flashSrc directory. This will be the preferred way to get the flash source from now on.
Bins for the Flash apps are now available here. I made some big updates to the photo app - pictures now slide around with some velocity after you let go and I fixed a bug with the two finger dragging behavior (when you let go with one finger it would snap to the wrong location before).
In other news, my employer (Archer Group) is having me build a table for them, so I'll be able to use a decent camera this time (we have ordered the FireFly MV from pt-grey research) and produce something that is a finished table, rather than a prototype. We are going for a table form factor.
The search for the perfect 'compliant surface' continues (someday we'll have to come up with a better name for it). Rockit (on the nui group forums) has reported some success with silicone rubber and mylar drafting sheets. I have a big piece of transparent silicone on order - I will let you know how well it works when I receive it. I'll be trying some of the rosco screen on the silicon rubber.
3 Comments:
Hi David,
I downloaded the sources for touchlib and the demos.
At the moment I'm trying to use the smoke demo. Compiling it in the "Debug" mode works very well (it's a very cool demo !).
but when I compile in "Release" mode, the application behaves in a strange way: instead of generating a smoke trail where the finger touches, it only draws a small mark (not the circle for the finger, but the beggining of the smoke trail) - but no dynamic smoke.
Did you try "Release" mode ?
Do you have an idea how to solve this problem ?
Thanks,
Gil (gilw@n-trig.com).
Yes, I've tried release mode. It works ok over here (haven't tried the latest version though). Generally problems with release vs debug builds are caused by uninitialized memory areas or other memory errors.
Whitenoise,
Great work on Touchlib.
I'm building a difuse illuminated multitouch surface, but can't get the "highpass" filter working in Touchlib.
Where in the config.xml file do I put the highpass section?
Cheers!
city_kid
Email: city_kid@eight33n.com
Post a Comment
<< Home