I am looking for a sensor module for my project. I would like to make a
contactless joystick, i.e. detecting up-down and left-right movements
made by a hand at close vicinity of the sensor(s), but without touching
it/them.
I initially thought about a square/diamond configuration (or even
triangle). By close vicinity, I am thinking a few cm (1 to 4 in).
Which type of sensor would you think would be more appropriate for that
kind of project? IR, sonar, Hall effect, other?
IR seems interesting but I fear that movement of other parts of the body
interfer with the hand movements. Pricewise and sizewise as well, IR
seems interesting, although I have close to no experience with other
types of sensors. Also, I am wondering how to deal with how to quit,
i.e. releasing the joystick. The final movement may give false positive.
I am not sure about prices either. Is there a big difference in using these different sensors?
What would you recommend?
Or perhaps such device already exists? I have not found one yet, though.
Many thanks in advance for your guidance.
Best regards,
There is not a big difference in the price of normal sensors.
If you want to measure the pH value of an aquarium or the dissolved oxygen in water, those are expensive.
Could you explain more what the project is.
With hall sensors a magnet is needed.
With IR sensors, it can be contactless.
Sonar might work, but perhaps a single finger is too small.
Radar is for security systems, and I don't know a radar sensor for very short distances.
What do you mean with not touching the sensor. Is applying the force a problem, or is it about hygiene. Should the sensor be covered by something that can be cleaned easily ?
You could combine an optical mouse sensor with a proximity sensor. The mouse sensor would detect movement of the hand and a proximity sensor would detect the presence of the hand. You would need to re-focus the mouse to give a greater distance between the mouse and the moving surface... the mouse would be focused at the contact plane.
A less precise option would be a group of IR phototransistors and one or more IR LEDs. By toggling the LEDs on and off you can distinguish a reflection from ambient light. The range in intensity values might be enough to detect proximity.
Koepel:
Could you explain more what the project is.
What do you mean with not touching the sensor. Is applying the force a problem, or is it about hygiene. Should the sensor be covered by something that can be cleaned easily ?
I would start a test with sonar.
I am looking for a sensor pad/pod/flat button fixed on a wall or a table for controlling devices that are "fixed" in the house: window blinds, TV channels and volumes, light brightness and colour, eventually oven temperature and cooking modes, etc (to control two parameters).
I envision the "contactless joystick" essentially because of hygiene. Let us say we are cooking, our hands are greasy, covered with spices or others and want to open the blinds. Instead of washing, soaping and drying the hands, we would pass the hand in front of the sensors upwards. To turn the blades of the window blinds, we would pass the hand horizonzally (left-right) in front of the sensors.
But it is not only a question of hygiene. I prefer to reduce the mechanical parts as much as possible, and have a flat surface.
I imagine the sensor pad/pod/flat button as a small square or round surface, the size of a fingernail or of a wrist watch (round), i.e. rather small. This sensor would control the mecanism of the blinds (already electric blinds). or other devices.
I want to NOT touch anything, as much as possible. Also, the sensor must be able to differentiate between the finger/hand movements and the movement of other parts of the body/face behind the finger/hand. I do not wish the blinds to open or close if we nod with our heads while the hand is moving sideways, or react in any way if people walk by, say at an arm length (30 cm or more / 1 foot or more). I want it to react at close distance, say a few centimeters up to 12 cm (approx. 1 to 4 in.).
I have seen some experimental devices, and some commercial devices that can recognise complexe finger gestures, such as crossing fingers, rotating fingers, double fingers gestures, etc.
What I am looking for/want to create is much simpler (and certainly as inexpensive as possible). I want only two movements: vertical and horizontal. Nothing complex.
Many thanks for your replies so far and looking forward to your next replies.
johnwasser:
You could combine an optical mouse sensor with a proximity sensor. The mouse sensor would detect movement of the hand and a proximity sensor would detect the presence of the hand. You would need to re-focus the mouse to give a greater distance between the mouse and the moving surface... the mouse would be focused at the contact plane.
Thank you for your reply.
I am very inexperienced with mouse sensors (and with most sensors). Is the beam a danger for the eye if directed into it? Sorry for the silly question, I have absolutely no idea what type of beam and power it has.
Also, if the mouse is fixed, what is the typical detection angle and sensitivity of such sensors, i.e. how much should one move the finger/hand to be detected properly (detection window)?
It sounds a nice system though. Sorry to ask again: do you any project using this mouse sensor in a similar way as mine, so that I can see how to proceed?
johnwasser:
A less precise option would be a group of IR phototransistors and one or more IR LEDs. By toggling the LEDs on and off you can distinguish a reflection from ambient light. The range in intensity values might be enough to detect proximity.
Thank you!
I actually use a certain number of PIR in my home for lighting purposes. but I guess PIR is too imprecise for this project. Sorry again for my ignorance: is the intensity and variation of ambient IR such that it is an interference to proper detection? In my ignorance, I thought that the IR emitted by the human finger/hand would be such that there would be a clear intensity gap between ambient and human generated IR. The other way around: is the IR LED necessary in indoor use? I am asking because I am trying to get the system as simple and inexpensive as possible. But I indeed know close to nothing in such systems.
Many thanks for your replies and for your next replies, if you please.
PanSer:
I am very inexperienced with mouse sensors (and with most sensors). Is the beam a danger for the eye if directed into it?
The mouse sensor is basically a low-resolution camera that senses optical flow in the image area. The 'beam' is a red LED use to illuminate the area being imaged.
PanSer:
Also, if the mouse is fixed, what is the typical detection angle and sensitivity of such sensors, i.e. how much should one move the finger/hand to be detected properly (detection window)?
In normal use the imaging area is a few mm in each dimension. When focused at a greater distance the area might be a cm or two. In normal use the mouses senses movements of hundredths or thousandths of an inch. At greater distances the resolution will be reduced so you may have to move your hand 1/100th of an inch per pixel.
PanSer:
It sounds a nice system though. Sorry to ask again: do you any project using this mouse sensor in a similar way as mine, so that I can see how to proceed?
Here is a project that uses an optical mouse sensor with an Arduino:
PanSer:
I actually use a certain number of PIR in my home for lighting purposes. but I guess PIR is too imprecise for this project.
That is right. A PIR sensor is designed to notify you when it detects a warm object moving in its area of detection. It doesn't say where the object is or which way it is moving so it would not be a good analog input device.
When you use a plain IR phototransistor for sensing, all you get is the amount of IR radiation hitting the sensor. This can be ambient IR from the sun, rook lighting, or other IR senders like a TV remote control. By having an IR LED you can detect presence of a nearby reflective object. To ignore ambient IR you can take readings with the IR LED off and on. If there is a reflective object nearby the IR LED bouncing off the object will cause those two readings to be different. The degree of difference will give you a very rough measure of distance to the object. You could do the same with visible light but the visible light flashing on and off might be annoying.
PanSer:
Sorry again for my ignorance: is the intensity and variation of ambient IR such that it is an interference to proper detection?
Yes.
PanSer:
In my ignorance, I thought that the IR emitted by the human finger/hand would be such that there would be a clear intensity gap between ambient and human generated IR.
Using a thermal imaging device to detect the motion of a warm body is MUCH more expensive than reflecting an IR LED off an object. They cost on the order of $100 each. You can get 100 IR phototransistors from Amazon for about $5.
The usage you describe might be better serviced with a voice recognition app on an android tablet. Use arduino & simple proximity switch so when you wave the hand near it, it gets the voice app attention (expect new command mode). Then android app sends back to arduino commands to turn on/off or up/down whatever devices you ordered and are connected to it.
The waving motion (that turns voice command on) can also momentarily turn off any noise producing devices (TV, music player, chainsaws) so they don't interfere with your clear vocal commands.
The tablet could display your command prior to sending it to arduino, to make sure it was correctly recognized.