Gestures are one of the most used Non-verbal communication system,other than the facial expressions,which are far less intentional,but comes naturally out of our faces.In this manner gestures are more easy to understand and readable at a glance as they are purposely generated as a main stream of communication since the time humans started to live in social groups.Not only among humans the gesture communication is present in many animal societies too.As we can understand and distinguish between different kinds of gestures just by observing them,is it possible for a electronic device to read the same as we do?Answer is Yes!
Like our eyes,the Camera on an Electronic device(In this specific case a camera on a mobile phone) acts as a input slot for these physical signals and then turn them into digital ones where a processor will decode it and recognize the meaning of it. Yet yield results are not on par with the biological human eye,but rather interesting to see how close they will get by the time.
Still these concepts may look just futuristic but we’ve already seen few amazing experimental work published,some may not directly link to this topic as it is,but others we can call them as previews of early steps of evolution.One such thing is the NinjaStrike game,in which the backside camera sensor detects the movement of hand across its visual path and then quickly translate that action into a digital signal.
Camera-based motion data is problematic to use for gesture recognition due to the low quality of the data, caused by a low sample rate combined with high noise.So as a solution most gesture detection systems use a gesture recognition algorithm that tries to solve this problem by using state machines, modeled from a gesture rule set, that parse the motion data and interpret the gesture the user has performed. Which is the core technology behind what Samsung has patented some time back.In this a particular gesture is identified by comparing it with a one stored in its gesture rule.So misinterpretations should be expected.Interference with surrounding environmental disturbances on the gesture input is much low than it’s for Voice input system but still the factors like light levels and other visuals barriers may stand in the way of this gesture readings.
Camera-based motion estimation was first proposed as positioning input for mobile phones by Rohs. And since then there are several Gesture input recognition algorithms developed and available but still non-of them made their way on to a public retail device,as a fully gesture controlled one.But in the recent past we saw some interesting and promising demonstrations involving touch less gestures where no contact with the device screen is done during the input.
Samir’s Hand Wave is one such innovative application which users the front camera sensor to detect input,
or using a smile to trigger a camera shutter
So all these things are indicating the gradual approach to a new level of input system that we will soon experience on our devices.Yet we are not there,but in close vicinity.