Watching your child on a swing is fun. Taking pictures for the family album is fun as well. Two things are tricky about it: getting the timing just right and getting your child to look at least near the camera when it goes off.
This project attempts to kill both birds with one stone by placing a small sensor board at the swing and subsequently trigger the camera by some kind of wireless link (radio or IR). The receiver on the camera triggers the camera and is also able to grab the attention of your child by sound and light effects.
Up to this point, the attention grabber and the swing-timing-thingy were two separate ideas but connecting the two was just too obvious.
Ideally, the timing-device should send a trigger to the attention-grabber a short time before the camera is triggered. The attention-grabber should at least switch off its light effects just before the picture is taken.
This leads to the question which mode of communication should be used between the parts.
My SLR camera (a rather dated Pentax K100d) offers two options for remote triggering: a 2,5mm jack which I previously used for my time-lapse-bot and an IR-remote.
Furthermore the attention-grabber needs some kind of input as well and is also supposed to work on its own, since I consider it useful in other scenarios.
The idea therefor is to establish a connection between timing-thingy and attention-grabber and use the attention-grabber to subsequently trigger the camera using the 2,5mm jack. Connection between the two components of the system could be established by e.g. 433 MHz radio or IR. I shall look into both options at a later point.
A few weeks later we discovered just how much fun swings can be. Taking a picture of a kid on a swing can be tricky. The best points for triggering the camera obviously are the highest points since the swing is not moving so there's no risk of motion blur.
Clearly, this is something that can be automated!
Since whatever system comes out of this project should be able to work on-the-go, and should require little setup and calibration time anything using a through-beam or reflective photoelectric sensor is ruled out.
A rotation rate sensor or an accelerometer seem to be a practical solution. Since the motion on a swing is typically sinusodial with a frequency in the order of 1 Hz it should be pretty easy to identify the turning points with a low-cost sensor and some microcontroller (no, I am not going to mention the A-word here.
I got the idea from an optometric examination of my infant daughter. The Auto-refractor used in the examination emitted a series of beeping noises and blinking lights to draw the attention of my kid towards the device. Beeing a nerdy dad's kid she usually looks towards LEDs and the like.
This and the fact that it is rather hard to get her to look in the direction of the camera got me thinking: would it be practical to have the lens of the camera surrounded by some blinking lights to get the attention?