Software Archive
Read-only legacy content
17061 Discussions

Capture common movement from a person

Rafael_Cronemberger
490 Views

Dear all, many thanks for your attention.

I'd like to ask you guys if it's possible to create the solution below using the Real Sense camera.

I would like to "teach" a very common movement to the camera which consists in a person picking up an object too many times per day. As I told, this is a common movement, and the person does this movement too many times per day.

Every time a person does this movement, I receive the event by the SDK.

Is it possible?

0 Kudos
1 Reply
MartyG
Honored Contributor III
490 Views

It should be very possible to do.  I have something like this in the RealSense-powered game I'm developing in the Unity game engine.  When the full-body avatar character touches an object with its virtual hand that has been programmed with a 'score' trigger, a score counter is increased by 1.  This score can be read in real-time on a pop-up window.

1.jpg

My own system is programmed to only count the first touch, with the first touch on various different objects adding to the counter as the player interacts with them.  But it would be no problem at all to leave out the turn-off instruction and have the counter incremented by one on every single touch of the object.

I couldn't tell from your description whether the object is in a game, or if you want to point the camera at a real-world environment and trigger the counter when a real-life object is picked up.  If it does need to be in the real world, I can think of two ways to do this.

1.  Place the camera at a certain distance from the RL object and set the camera to have a "Real World Box" (an imaginary box superimposed over the real world environment) that the object is located within.  The camera would only start tracking motion when a recognizable feature such as a face or a human hand entered the boundaries of that box.

2.  Or use a feature called Scene Perception to have the camera "cut out" a real-life object from the image of the environment that it is looking at and treat that highlighted object as an interactive object that can have effects associated with it.

0 Kudos
Reply