Gesture to speech is a project specially designed to help the differently-abled people! As the name suggests, this project will enable the people to convert their gestures to speech so that they can convey their messages.
“Must be sign language” Is this what you are thinking? Well, it is not sign language, the gestures applicable here are just hand sway from one side to the other!
Let us see, what are the master components used in this beneficial device prototype?
It’s the APDS sensor!
What is the APDS sensor?
It is an all-in-one sensor which comprises of:
Proximity sensor
Gesture sensor
Color sensor
Ambient light sensor
Gyro sensor
Accelerometer sensor
Among these, we are making use of the Gesture sensor. The gesture sensor can detect 6 hand movements namely - Right, Left, Up, Down, Far and Near.
Let me give one more light on this point, keep the sensor in a horizontal position and make sure the port is toward the left. Moving your hand from left to right would mean ‘right’, the other way would mean ‘left’. From down to up would mean ‘up’, the other way would mean ‘down’. And finally, moving your hand close to the apds sensor and bringing it straight up would mean ‘far’ and from far bringing it close would mean ‘near’.
TOP TIP: For the best results and better readings, make sure to use only two fingers and not all the five while swaying!
There are 26 alphabets, how can I cover all of them with just 6 gestures! The solution to that is Math, simple division. We can categorize the alphabets in groups, those in a particular group will be decoded using the same gesture, let’s take an example,
A B C are to be conveyed using only ‘right’ gestures.
1 sway to the right would mean A,
2 sways would mean B,
3 sways would mean C.
Now, what if the person does a fourth sway? Well, we can create a loop and send the fourth sway back to number 1(Using a separate if loop: if (sway > 3 then set sway = 1)). So now, the problem is solved!
We can’t categorize them into 6 groups but 4, why? Let’s take another life example.
You go to the store to buy some clothes. You see a stack and SHUFFLE through it. Once you get the desired cloth (let’s say the 4th one), you pick it up and go to the billing section.
The same way, we must keep a separate gesture for ‘enter’. To pick out the desired alphabet from the rest.
The other must be kept for ‘space’ to differentiate one word from the other.
VOILA!
We have now covered all the alphabets and they can be decoded using the above gestures. This is the main principle on which the ‘Gesture to Speech’ works.
Also, there are two limit switches present, one for ‘Backspace’ the other for ‘Speak out’. They can easily be felt by a person who can’t see. Once the person thinks that he or she is done, he/she can click on the assigned limit switch so that it can speak it out!
Do you like to build one? Connect us - www.qtpi.in
Credit: Article written by Srija Ghosh