Sensor to track moving person?
ionymous
Posts: 29
I'm brainstorming a project.
It's a cable cam to follow and film me as I cycle through the woods. The general concept has been implemented by others many times. A carriage with an attached camera rides on pulleys on a cable suspended between two trees. Ideally the camera keeps the subject centered in the video. Sometimes this is implemented with a motor and a remote control, to be controlled be a third party. Sometimes the cable is angled and the carriage moves by gravity.
I would like to implement an autonomous motorized version. This eliminates the need for a third party, and allows uphill movement.
A microcontroller will use some kind of sensors to determine if it needs to speed up or slow down a motor to keep me centered in view. I assume it won't be analyzing video because that would be too costly. Correct me if I'm wrong.
I will be using a GoPro HD camera, which has a very wide angle. It actually looks best only about 4' away. But I'd like the ability to get about 20' away and still track me. I will generally be doing more technical riding and hill climbing. So generally I'll be moving at about 5mph or less, but I may get up to 10mph occasionally.
At first I thought a wiimote IR camera and IR leds could be used. I would wear the IR leds, and the wii camera would track my IR blob. Unfortunately sunlight confuses the wiimote too much.
I then started looking at ultrasonic sensors. I could wear a transmitter that pulses an ultrasonic signal. The carriage could have two ultrasonic receivers mounted on opposite ends, like ears. Based on time of arrival, the microcontroller could determine which way it needs to move.
Any thoughts on this?
Will ultrasonic work?
What would be better?
Visible light would be seen in the video... so that's not preferred.
RF?
Something else?
Thanks for any suggestions.
Ion
It's a cable cam to follow and film me as I cycle through the woods. The general concept has been implemented by others many times. A carriage with an attached camera rides on pulleys on a cable suspended between two trees. Ideally the camera keeps the subject centered in the video. Sometimes this is implemented with a motor and a remote control, to be controlled be a third party. Sometimes the cable is angled and the carriage moves by gravity.
I would like to implement an autonomous motorized version. This eliminates the need for a third party, and allows uphill movement.
A microcontroller will use some kind of sensors to determine if it needs to speed up or slow down a motor to keep me centered in view. I assume it won't be analyzing video because that would be too costly. Correct me if I'm wrong.
I will be using a GoPro HD camera, which has a very wide angle. It actually looks best only about 4' away. But I'd like the ability to get about 20' away and still track me. I will generally be doing more technical riding and hill climbing. So generally I'll be moving at about 5mph or less, but I may get up to 10mph occasionally.
At first I thought a wiimote IR camera and IR leds could be used. I would wear the IR leds, and the wii camera would track my IR blob. Unfortunately sunlight confuses the wiimote too much.
I then started looking at ultrasonic sensors. I could wear a transmitter that pulses an ultrasonic signal. The carriage could have two ultrasonic receivers mounted on opposite ends, like ears. Based on time of arrival, the microcontroller could determine which way it needs to move.
Any thoughts on this?
Will ultrasonic work?
What would be better?
Visible light would be seen in the video... so that's not preferred.
RF?
Something else?
Thanks for any suggestions.
Ion
Comments
It comes from a Circuit Cellar article in Byte (collected in Ciarcia's Circuit Cellar III):
http://books.google.com/books?id=zAsN0XCHbFsC&pg=PA42&lpg=PA42&dq=ciarcia+byte+parabolic&source=bl&ots=br9TduIhF6&sig=t4wLtDMCk2Ws9vt7tC15YkLSPRU&hl=en&ei=SRmSTKzDN8WBlAfrkrikCg&sa=X&oi=book_result&ct=result&resnum=1&ved=0CBIQ6AEwAA#v=onepage&q&f=false
I probably have the book around somewhere. I used to have the original issues, but they got accidentally thrown out during a move.
Robert
IR won't work outside.
Would ultrasonic be worth testing? Wearing an ultrasonic beacon, it will be bouncing around with me. Is that a concern? Vibrations on the trolley?
With a beacon (instead of echo) can I get more distance?
Any other ideas?
Robert