Click here to Skip to main content
15,907,687 members
Please Sign up or sign in to vote.
0.00/5 (No votes)
I already have a subsystem for my appliances. The only thing I need is to trigger it to turn on using my camera. I hope someone could give me a C++ code that would make my camera able to detect human and trigger my system to turn on. I am using Zigbee through rs232 and i was hoping that the code can send data to my zigbee.
Posted
Comments
Joan M 4-Jul-11 12:22pm    
You should give us more details: which visual event do you want to trigger the information transmission (camera part)? we know the physical medium that you will use to connect to your appliances (RS232), but not the software layer of it... you've spoken about a zigbee... does this has a description of the data it expects to receive and the ways it has to receive it?
As far as I can see it is a two parts question: image side and transmission side... could you separate the questions and give more details on what do you need? ;)
Mohibur Rashid 12-Jul-11 22:38pm    
Wow you are looking for an intelligent software that can detect human presence on a forum..

Good luck with getting it free. (of course we have got linux as a gift :P )
Mohibur Rashid 12-Jul-11 22:45pm    
Anyway, My comment was little harsh before...

But I understand that you want to start saving video after you detect some ones presence in certain room.

For a still camera(that does not move his head) you can do simple image comparing, change i mean, if image change then you will start storing image.

But figuring out change or movement wont be easy(i have no idea about image processing.). because in simple image processing that will judge bit by bit will trigger even if light changes. that is if the day light start to deem it camera will think a that a movement generated.

if the camera is moving camera then everything is moving.

So, its not a easy job for sure

Just a suggestion here: do you really want to do this with image recognition? It certainly isn't the simplest approach, and it may well produce a lot of errors (e.g. dependency on lighting conditions makes a dog looks like a person, person look like a lamp and so on). Following the KISS principle, perhaps you can achieve your goals with a much simpler setup using an RFID tag, bluetooth polling, or alarm system motion detector. If you are more specific about your setup you will likely get more useful answers.

I will be most interested to see how your image recognition works out should you pursue that however!

good luck!
 
Share this answer
 
Thanks for the time to reply and read my question. Apparently I was able to make the image using the camera trigger my program in my mcu. I just used libserial in ubuntu to communicate with my max232 and my xbees... If your interested on the code just mail me.
 
Share this answer
 

This content, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)



CodeProject, 20 Bay Street, 11th Floor Toronto, Ontario, Canada M5J 2N8 +1 (416) 849-8900