How motion detectors work

HAYDEN WALLES
Last updated 02:46, December 3 2013
HOLD IT RIGHT ERE: A look at the Xbox One's new Kinect motion sensor.

HOLD IT RIGHT ERE: A look at the Xbox One's new Kinect motion sensor.

Building a machine that can detect motion is potentially a very complicated business.

One way to go about it is to attach a camera to a computer and devise sophisticated software which isolates the objects you're interested in, keeps track of them, and lets you know when any of them move.

That's how the Kinect, the sensor for the Xbox game console, does it.

If you need to extract details about the motion and use it to control a console then the complicated software is really your only way out.

But there are many situations where all you want to know is whether something moved or not, and simply detecting any motion is a much easier problem to solve.

There are other ways to detect motion besides literally seeing it.

Motion has other side effects that are easier for machines to detect.

For instance there's the Doppler effect, the phenomenon that makes the pitch of an ambulance siren rise as it comes towards you and drop as it speeds away.

The Doppler effect applies not just to sound but any kind of waves, including light and radio waves, and in the 1920s Edwin Hubble used the Doppler effect to show that the universe is expanding - motion detection on a cosmic scale.

The Doppler effect is not limited to waves emitted by a moving object, though, it also works for echoes, which back on Earth lets us build a motion detector.

Ad Feedback

The motion detector simply broadcasts a constant tone (using ultrasound or even microwaves) and listens for the echo. If the echo is lower or higher than the emitted tone then something is moving in the environment.

But probably the simplest kind of motion detector, and probably the most widespread, does its work by taking the temperature. Passive infrared (PIR) sensors are best at detecting moving people, and you'll commonly see them employed in burglar alarms and security lights.

The hotter something is the more infrared it emits and, since we are generally warmer than our surroundings, humans stick out like a sore thumb in infrared.

At the heart of a PIR sensor is a small piece of semiconducting material that electronically measures the total infrared shining on it.

A lens focuses all the infrared from the desired field of view on to the sensor so that the sensor effectively measures the average temperature of its surroundings.

When that average temperature changes - as when someone walks in front of the sensor - an attached circuit detects the sudden change and signals that motion has been detected.

Actually PIR sensors typically have two halves, one covering each side of the field of view.

Motion is detected only if the sensors fire one after the other, which happens when someone crosses the field. This eliminates false positives from things that affect the whole field simultaneously, like temperature changes, flashes of light and other distractions.

It's not perfect but it is surprisingly effective for such a simple system.

Our fast changing world may seem to get more complicated every day but - as with many things - when it comes to actually detecting motion it is often the simplest solution that works best.

 - Stuff.co.nz

Comments

Ad Feedback