Creating Personal Radar: Software or Sensors?

[jhe1984]

Would it be possible for me to make a sort-of personal radar program for my laptop which would hook up to my comp via USB or something and which would display objects on my screen like people moving, etc?

Actually, it's probably more like a visual motion detection software program, but nevertheless is it possible?

What would the hardest part be: designing the software or figuring out how to attach various motion sensors to my laptop?

Thanks yall.

(PS - I hope this is the correct forum for such a question)

 

[Danger]

If you're looking for the kind of effect that I think you are (like you see on TV shows such as Alias or Nikita), it's not going to happen. Those are FX graphics. You could approximate it, but you'd need a vast number of sensors and very intricate data manipulation. You could look into some industrial security companies and see if they have something available, but you probably couldn't afford the top end stuff. If it's just a matter of wanting to keep track of people, I'd recommend sticking with video surveillance. Of course, a lot depends upon how large an area you want to cover.

 

[jhe1984]

Yeah I haven't seen those shows but it sounds a lot like what I had in mind, low tech sensors etc combined with some sort of graphics.

Basically, I was just wondering whether it'd be possible to sit in a place like Starbucks with my laptop and as people walked by within the circumference of the motion sensor or whatnot, I'd be able to see the corresponding motion on a graphics display on my computer. I don't think it'd have any practical applications; I was just wondering if it'd be possible.

It does seem like you would be able to take a motion detector and possibly the attached CPU that translates it and be able to transmit that info graphically on a computer.

Any thoughts or cost estimates?

 

[Pengwuino]

The only thing i can imagine that does anything like that is one of DARPA's projects... one of the features is where you actually have commanders with maps with little dots showing where people are. The problem is that ... its GPS... not radar...

 

[Danger]

Another thing to consider, jhe, is that you would also have to have complete mapping of the surroundings in order to properly reference your dots. In the case that you just mentioned, a floor plan of Starbucks would have to be in your database.

 

[Longstreet]

That would be an interesting project. My first thought was to make some kind of scanning rangefinder, and track changes in the distances recorded. But I imagine the accuracy would be too poor to create a meaninfull graph.

 

[Pengwuino]

That would be an interesting project. My first thought was to make some kind of scanning rangefinder, and track changes in the distances recorded. But I imagine the accuracy would be too poor to create a meaninfull graph.

and you really wouldn't know whos who, what's a person and what's a chair, etc etc.

 

[jhe1984]

True, but if you could combine a range finding app with a motion detection app, it might be possible to pick out a moving person from a chair, etc. But again that's only for moving objects.

The point about needing a reference map (like the floorplan of a Starbucks) is one I hadn't considered. While it wouldn't be possible to say "bob is walking to the counter" w/o a floorplan, wouldn't it still be possible to say "object A is moving at a certain rate of speed in a certain direction" simply by having the comp+sensors be stationary??

[EDIT: Also, I wonder if the lack of a floorplan could be overcome to a large extent by having your comp+sensors in the same location for a large period of time - overnight would be best, but I imagine several hours inside would work as well - and that way you'd be able to develop a pretty accurate picture of where static objects are versus dynamic objects like people.]

 

[Danger]

You're right about not needing a map. As long as you're satisfied with having your dots relative to each other and to your monitor, you'll be fine without it.
As for differentiating between people and inanimate objects, I would recommend incorporating infrared scanners. A moving heat source is then the only thing that will trigger it. If you can get your system to 'lock on' to one, then it should continue to register if it stops moving. That would be really tricky, though. Something like a steam table at a smorg or the urns at your Starbucks would stay on screen because of people in its proximity setting the lock function.
I think that you have your work cut out for you. Good luck. If I think of anything else, I'll let you know.

 

[jhe1984]

Thanks for all the help!

 

[dimensionless]

You could also do it, at least to some extent, with high frequency sound waves. It would be difficult though. I can't say what part would give you the most trouble.

 

[Anony-mouse]

they have had handheld radars for years

I saw one in janes (its a group of publications that covers everything military from small arms to battleships and everything in between) and it was over 10 years ago.
it was a hand held device with a screen similar to the dector from aliens but work with radar instead of air vibrations and was real as opposed to a movie prop. I have not been able to find it online since though. I saw something about how to build one that worked on IR instead of RF. well if you find one please post it.

 

[Q_Goest]

Hi jhe. Not sure if this is of interest, but check out what this guy does with a Wii remote:

http://www.ted.com/talks/view/id/245

 

[mgb_phys]

An ultrasonic rangefinder plus a stepper motor would be easiest
eg. http://www.robotshop.ca/home/suppliers/devantech-en/srf02-ultrasonic-range-finder.html

Or if you want to go a bit more mil-spec http://www.mdl.co.uk/laser_modules/laserace-slm/index.html

 

[Danger]

Q, that's an awesome video. It's starting to give me some ideas... :uhh:

 

[russ_watters]

That would be an interesting project. My first thought was to make some kind of scanning rangefinder, and track changes in the distances recorded. But I imagine the accuracy would be too poor to create a meaninfull graph.

There is a recent thread on the subject, complete with screenshots from a device that does exactly this: https://www.physicsforums.com/showthread.php?t=224021&highlight=laser+model

Here's the device: http://www.cc.gatech.edu/ai/robot-lab/research/3d/
It is truly extrordinary.

There are actual 'radar on a chip' devices (google it) but they are probably pretty expensive.