I want to create my first kinect application which is will be a grid gallery where the user can click to enlarge and read description. All the data is stored in a xml file. I've never used kinect so I would like some advice on the best way to achieve it. I've made similar applicatios with as3, it is possible, for example, to replace the mouse input with a gesture?
Thanks.
If you prefer working in ActionScript, there is AIRKinect. It features a complete Kinect data access layer, but nothing beyond that. You will have to implement gesture recognition yourself.
Performance-wise, doing Kinect stuff with AIR is a bad idea from my experience. Using C# with the official Kinect for Windows SDK is much faster and, in my opinion, easier to use. Some tutorials and documentation can be found here.
This SDK doesn't provide gesture recognition out of the box either, so you might want to give Kinect DTW a try. Others are available - Google will help.
Your question was basically if one could replace mouse input with a gesture. You certainly can. Just cause whatever action you want when a certain gesture is performed.
As for selecting photos, in case of C#/WPF, you can modify the mouse position while it's inside the window. You can use one of the hand joints from skeleton tracking to calculate an appropriate mouse position.
For the actual act of selection (which would usually be the click) I don't recommend gestures. Commercial Kinect applications prefer the approach of a visual timer counting down as soon as hovering over a button, and activating it as soon as time's up. This approach avoid the problem of accidently moving your hand to a different position while performing a gesture.
You asked a really broad question, so all I can do for now is give you some pointers. With the resources linked in this answer, you should be armed and ready to get started on your own. Best of luck!
