Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Person counter for doorway

Status
Not open for further replies.
Hi again.

Regarding the Infra red TX & RX. I am having difficulty calibrating the reflected 'range' downward for narrow spaces. By varying the current to the Tx LEDS via a trimmer I am able to get it close but not to within a couple inches accuracy.
The approach I'd like to try is to double duty the 38Khz carrier (as it is driven by the PIC's CCP module) as a modulated PWM to create a voltage reference for a transistorized current limiter. That way the uC can do its own auto 'gain' control for the Tx I.R. LEDs for a given doorway by adjusting the LED current drive until just below the receiver trigger threshold.

Sounds reasonable?
 
I'm just asking for suggestions as to the approach. How does buy & sell get into it?
Haha, I asked for various reasons. If its not commercial and just a fun project then fair enough. If not, then you could always buy something off the shelf.
 
Last edited:
I have designed the most reliable method using transmission interruption using an array of 5mm Sharp or equiv IR Emitters
I chose a number of emitters placed in critical locations to detect the body without arm paths that create multiple counts. Thus several vertically spaced emitters are needed detect the chest position. Anti-bounce detection code was created to ignore false triggers. This was done using paired emitters spaced apart by a few cm at the same height and other pairs at different heights, so that as long as one pair was detected in successive events of blocked light with the same direction , then the direction can be determined and the object with redundancy. Time filter allows a wide range of walking speeds but slow enough to reject small objects (arm hand..book blocking the path that causes false triggers.

Our code was implemented in a couple hundreds bytes on a PIC. ( proprietary)
I used Sharp IR LED and Rx modulated carrier to send unique 8 bit bursts Mux'd to each LED sent from a PIC.

This way each sender position could be decoded in TDM fashion error free or blocked by the path. and direction and reliable object passing results.

In the end, we used it for BOB detection on shopping carts or Bottom of Basket security aide for cashier clerks. It could detect a case of beer in the bottom of American shopping carts... or a resistor wire crossing the path.
Accurate and reliable. Simple yet complex in Cross-correlation time filtered TDM array mapping with 7 LEDs.. perhaps only 4 needed here.
 
Thanks.. Sharp mastered the quality on SFH LEDs and IRDA Rx receivers now cross-licensed by others. They are quite reliable for 1 meter communication and useful for 3ft people counters.

But the key is path interruption . The sender is autonomous so no connection required on both sides but they do need to be aligned so the beam covers all the recessed holes of the IRDA detector. Since the Rx has a daylight blocking tinted filter, the flicker of FLs has no effect.
 
Probably the only way to keep track of each person is give everyone that enters the building a bar code tag they are required to wear so a laser can scan it as they pass through the door.
 
Last edited:
One could use a simple laser beam with a diode laser at waist height and use 200 ms retriggerable one shot to avoid arms. But using an IRDA port with AGC and detecting communication errors with a repeating simple data pattern unique from each LED sent used for leading edge detection of blocking light for person going in or out is far more reliable from my experience , lower power and far less sensitive to alignment with a 10cm beam pattern @ 1 meter rather than a pulsed laser. Although you use a large diameter diffused cover to receive light in a wider zone. then interruption of each beam is like below with false edges even with hystereis for going in and out.
WHen both are high ( Light blocked) and sequence is same when Light is On it is a confirmed person in one direction and same for other direction.

A one shot for each logic level detected from optical sensor with variable retrigger time filters out the glitches.

______|----~--|-|-|-|______|-||---------~-|--|-|-|________
___|||-------~-||_____________|||----~------------|-|-||-_________

Narrow Optical path AGC is critical cover wide variations of receiver gain. Perhaps a remote control beam could be used with alternating senders. and common receiver.
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top