Motion controlled camera

Forum 2005-2010 (read only) General Exhibition
1

Hey!

I'm thinking of making a motion controlled camera. The camera itself shold not be motion controlled, but the rack that holds the camera should be motion controlled.

I've thought of the following setup:

I use 2 pir detectors that will give me analog outputs. The sensors would be placed like this:

THIS SIDE IS FACING THE AREA YOU WANT TO MONITOR
|
|
|
|
|
|
|
|

PIR SENSOR 1 PIR SENSOR 2
THIS SIDE IS MOUNTED ON THE WALL

Each sensor would be placed in the corner between the horizontal and vertical wall on the rack. The sensors would then cover 90 degrees each.
If I mount a fresnel lense on the sensors, that would make the sensors measure motion 90 degrees.

I am gonna use PIR sensors that measures infrared light between 4 - 15 micrometer. Human heat radiations sends out a infrared light on those wavelengts.

The thing I'm gonna try to do is to compare the readings from both of the sensors until they are alike (or as good as alike as they can get). And depending on the output from the sensors, the motor that controls the rack would turn the rack.

If sensor 2 detects motion (heat) the motor will turn the rack until the output from sensor 1 and 2 are alike. And they will be alike if the vertical wall on the drawing is pointing directly at the heat source.

I would like to hear your opinion on this project. And I'm thankful to every input I can get.

The sensor I'm planning to use is this:PIR D203S Datasheet. Do you think this one is suitable for my project?

  • I know that I can only control the camerarack horizontaly.
  • I know that if the heat source is above or below the camera, the sensors would still sense the heat and move the rack.
  • I know that if there are 2 people in the sensing area, I would get a problem.
2

It sounds workable; you are basically making what is known as a "sun tracker" circuit, but using IR detectors instead of visible light detectors. Instead of a large shield (well, I don't know what size you planned, actually), you could mount each sensor at the end of an IR opaque tube. If you used four sensors, you could also (in theory) track in the vertical axis as well.

As you note, with more than one person in the view of the sensors things won't work as expected (what you might do there, though, is seek and track to the "highest average" temperature/IR level - in theory, that would position the camera in the middle of a group of people, or near the biggest group and the smaller group, and for a single person, on that person alone).

:slight_smile:

3

Hey there!
I am brand new to Arduino, and just stumbled over your post :slight_smile:

Just wanted to share some thoughts that ive done over the same idea. You could look into the software called Isadora, it offers different tracking possibilities, IR aswell. I read around their forum and there are a few there working around with tracking and servos, (expensive software though, but they offer a trial)
You could find a solution through straight camera color/IR tracking, but maybe its not precise enough

Maybe this is not that helpfull, but I hope you will find a solution :slight_smile:

best regards

4

Hey Grinder!

Nice idea, added to my veery long to-do list.
Did you get to try it in practice yet?

Cheers, fatzo