GPS and the pseudolites cited don't use signal strength to determine distance but, rather, phase and timing differences in the received signals. Nonetheless, the RSSI approach may still provide a rough approximation of distance, all other things being equal. But there's the rub. Received signal strength will also depend on what the signal has to pass through to get to each receiver. For example, if the person is wearing the transmitter on his chest, like a name badge, the signal strength may depend on which direction he's turned relative to the receiver. If the space is not empty, any obstacles could also absorb the transmitted signal. These influences could be minimized by mounting the receivers very high, but then precision would suffer, since the differential distances from each one to the transmitter would be minimal.
Here is a paper on the subject of indoor location, and here is a link to one of the products mentioned.
How about hats with IR transmitters facing up. Each person's hat transmitts a unique code and it's altitude above the floow. You could then suspend a grid of IR recievers above the floor each looking straight down and reporting the altidude of any hats in it's sector.
It's been a while so I don't know if you've come up with an idea or not, but I've used sonar to track the front of a moving object and it's pretty easy, and using that you can calculate the angle of the servo and the distance to the person and put it through a trig function to get the x,y coordinates you want. To track multiple people you could put a sonar on the other side, and to prevent it from confusing targets it could search for differences in distance to prevent it from jumping to the wrong person.
Comments
-Phil
I was thinking I could use 3-6 transceivers (for instance: http://www.parallax.com/Store/Accessories/CommunicationRF/tabid/161/CategoryID/36/List/0/Level/a/ProductID/582/Default.aspx?SortField=ProductName,ProductName) as the pseudolites. Then have the object being tracked carry a transmitter. Sampling the RSSI line on each receiver (pseudolite) would provide absolute distance measures.
Is this the right line of thinking?
Here is a paper on the subject of indoor location, and here is a link to one of the products mentioned.
Good luck with your project!
-Phil
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Searider
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
PG
www.robotshop.ca/hagisonic-hg-b40c-ultrasonic-sensor-1.html
www.acroname.com/robotics/parts/R11-6500.html
Each one can detect a single person, if placed in the center overhead you may only need 3 to detect one person in the area of each one with pan/tilt, if you have it on the edges it will take six but it won't have to have a tilt mechanism like the overhead one. If you don't mind not knowing the constant coordinates of people in the area you could have it rotate constantly and display the position of any object not matching with the walls.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
PG
If its in the 10s of thousands of dollars.
Check out ultra wide band rfid setups.
I have seen this type of system set up outdoors. It didn't work great for us, but might fit your need.
200'+ range, 1-50 Hz update, 30 cm resolution.
If your interested pm me and i will try to find the name of the company they used.