DETECTION USING INFRED AND AUDIO SENSOR

You need to provide more details. Your question is like asking to paint a room, only we don't know how big the room is or what color of paint you want.

Okay i am doing a mini project that detect elephants using infrared sensor,the infrared consists of the transmitter and receiver,if an elephant walks into the line of sight,the information will be send to the microcontroller,the microcontroller will then interpret the amount of time taken when there was no clear line of sight and compare it with the average walking speed of an elephant,then the sound sensor too will capture the infrasonic sounds unique to an elephant and send it to the micro-controller,if this two inputs match or fall within a certain range,the controller will conclude its an elephant and send an output of buzzer sound.

The elephants walking thru a narrow path / chute ?

Can you program, prototype circuits, etc.....?


Regards, Dana.

i am working on making the prototype and circuits with codes but as you can see i am strggling beacuse there is little information and i am not an expert,just an undergraduate

Have you considered using passive Infra Red sensing? See https://www.futurlec.com/PIR_Sensors.shtml

Mike.
Edit, you may also be able to detect infrasound.
Edit2, Sorry, I see you know about infrasound.

Of course if elephants not in single file as they pass that poses detection
problems. Can elephants be tagged ? Say with an IR mirror ?

If not code/design must be intelligent enough to handle 2 elephants walking
side by side.

Or stipulate as part of the solution their only access is walking single file
past detector.

You can put detector at ground level, but that complicates leg positions causing
false counts. Or at shoulder level, but then that causes issues detecting baby
elephants.......


Regards, Dana.

IF you look for Vishay IR Rx used on remote controls . You can send carriers of different frequencies e.g. (16 kHz) with different pattern codes to uniquely identify Tx patterns. THe IR chip will have black daylight blocking filter lens and AGC to automate gain and levels for detection of binary data. Otherwise you can choose a fixed threshold and open loop, but this has lower SNR.

Tony Stewart said:

IF you look for Vishay IR Rx used on remote controls . You can send carriers of different frequencies e.g. (16 kHz) with different pattern codes to uniquely identify Tx patterns. THe IR chip will have black daylight blocking filter lens and AGC to automate gain and levels for detection of binary data. Otherwise you can choose a fixed threshold and open loop, but this has lower SNR.

Click to expand...

You've just got to train the elephants to carry a remote control, and press the button when they go past

Normal frequencies are 36, 38 and 40KHz - although the individual IC's seem to make little difference which type you use.

Nigel Goodwin said:

You've just got to train the elephants to carry a remote control, and press the button when they go past

Normal frequencies are 36, 38 and 40KHz - although the individual IC's seem to make little difference which type you use.

Click to expand...

Changing the assumption is not implied, which was,

"if an elephant walks into the line of sight.

So the Remote control is just fixed and the Detector is fixed to create the fixed path with a deep aperture so there can be no reflections strong enough to be false counted.

A fixed Tx IR LED to fixed Rx IC recessed to reduce the aperture beamwidth is what was assumed.

A path length up to 30m is possible with 8 to 10 deg IR Tx (emitters) with > 100 mA peak avg. pulses.

But you were correct the carrier frequency is higher from 30 to 60 kHz https://www.vishay.com/en/ir-receiver-modules/

Be aware if using a 38kHz modulated beam, it needs to be intermittent or the AGC of the receiver steps in and it (the receiver) stops seeing it.

Mike.

All the AGC IR Rx require a minimum number of carrier cycles and depending on the 4 models of AGC a maximum cycle count and they all have a minimum off cycle count of 10 cycles of say 38kHz (the most common channel.) The IC has a bandpass filter with a Q = 8 to 10 for option AGC=5.

You have a lot to learn on communication circuits, bit error rate factors, and defining system level specs. such as rumble audio noise levels and type of mic that detects this low frequency, methods of detecting "line of sight" baseband modulation signals and methods of rejecting noise. Defining SNR required for 0 false positives and true negatives. Defining optical beamwidth of line of sight and reflectance paths and how to detect false reflectance paths. How to define combination logic vs sequential logic etc. etc etc

Your flow chart has many flaws. Find out how Remote controls work with command acceptance and path loss. Characterize object patterns of motion at various speeds and directions. Use more emitters & detectors to detect moving object to improve detection.


I have done this before using the bottom of a shopping cart to detect objects above the bottom tube rail. It was very accurate and could detect a resistor wire blocking the path of a 10 deg IR LED at 1m gap. It detected the shape of an empty shopping cart precisely with 7 paths. Your problem is harder with variable size moving objects. Mine could detect long skirts as a false object but detect any cart accurately with rapid scanning different data words for each LED Emitter paired to matching Rx. There was no communication between Rx to Tx as in your flow chart. A PIC uC drove each side independently.

good luck.

I have read the working principle of your circuit. Never mind, I doubt that IR sensors will not always be accurate for elephant detection. The average walking speed of adult and baby elephants will not be the same.

I think the better option is to use a Raspberry Pi and train it.