Hello guys,

I live in a condo with 3 diferente garage areas. Each area is independent from the other and in different places. So it causes a little (but sometimes inconvenient) problem: Once you enter one area your don't know if there's free space to parking, so you have to leave and proceed to the other two until you find a place.

I'd like to receive suggestions on how can I build a system that identifies how many spaces are free and then show a signal in the entrance of the respective garage.

Please, if possible be very specific. I study electrical engineer but that is going to be my first real project (i'm still in the beggining) and I'm quite lost without knowing how to start, but I think if I accomplish this task it will be important for me.

Thanks

 

First question:

Will they allow you to install anything in their building?



Unless it's very few parking lots you will need to count how many vehicles enter and leave the garages. This will not be so difficult to achieve with direction detecting light barriers which you can build by yourself or buy off-the-shelf.

You will then need something to display the number of free parking lots or simply if there is any free left or not. This display you could install at the garage's entrance. The other option is to send this information wirelessly to a receiver in your car.

Please answer the following questions:
1. How many parking lots are there for every garage?
2. Are there multiple entry/exit points?
3. Are you able to program a microcontroller? (not absolutely necessary)
4. Do you want to use off the shelf devices or build your own circuit/PCB etc.?

This guy built something that could be used, after some adjustments...

 

The trouble with counting the entries and the exits is that if there's ever a mistake, it gets carried along forever. How reliable will the sensing and counting be?

You could paint a big X on the ground at each space, and have TV cameras and video software to count the number of X's left visible (unless someone has one painted on their car roof, ha ha). But that might add up to a lot of hardware.

Edited to say that maybe the vision software can be trained to recognize the background to each space, so painted marks wouldn't be necessary. That means getting a consistent result regardless of the lighting conditions, though.

 

The trouble with counting the entries and the exits is that if there's ever a mistake, it gets carried along forever. How reliable will the sensing and counting be?

I thought about that... It would need to be quite reliable to be useful.I think it could work for cars much easier than for people. Another option would be induction loops, but unless it's the OP's building I don't think he will be able to implement those.

 

If you can get a camera or cameras to view the parking lot that would be the way to go. It shouldn't be too difficult to do image recognition. Matlab has libraries to do this.

 

If you can get a camera or cameras to view the parking lot that would be the way to go. It shouldn't be too difficult to do image recognition. Matlab has libraries to do this.

For a "first project" this sounds quite difficult to me.

I'll be back soon to ask you about that image recognition for my projects.

 

Well, actually, image recognition with Matlab is a rather fun project that does not require any hardware knowledge and not much math. All the libraries are there. You just connect a web cam or any USB camera and away you go. Some background knowledge of FFT helps but not essential.

Start off with a tray of nuts and bolts and see if you can count the different components.

 

fast food drive throughs use a hose capped off at one end, and a piston on the other side that stikes a bell (or at least they used to). Take this design, replace the piston with an adjustable pneumatic pressure switch. calibrate the pressure switch to switch when a car drives over it but not when a kid jumps on it. 2 switches within a prescribed window of time time = a car. too fast (car would have to be going 100mph) and its not a car; too slow (car would have to stop straddling the hose for 10 minutes) and its not a car. this further lessens the likelihood of false triggers. put one on the entrance and one on the exit. entrance adds, exit subtracts. use a microcontroller.