Print PDF
Favorite
Email
Step 4Learn to semaphore
Now the fun part. You can learn to semaphore by looking at the dialpad and keyboard. This shows you the positions you have to adopt in order to choose the letters and numbers you want.
Once the application starts up (or if the background seen by the camera changes significantly) it will take a few seconds to calibrate. During this time it will be confused and flash yellow and blue squares.
Once this has finished, you can move your yellow and blue 'flags' and the yellow and blue squares should faithfully follow your position. With the squares in the right position, just hold still and after about half a second, it will type the letter for you. If you need to repeat the same letter, this takes around two seconds.
If you make a mistake, signal 'DELETE" and it will remove the last character. When you've finished one stage of the message (e.g. dialling, writing the 'from', writing the 'subject' you can signal 'NEXT' to move on to the next stage. After you've filled out the body of the message, signalling 'NEXT" will send it, and then return you to the beginning again.
Semaphore is still taught and used in the Royal Canadian Navy, so you may even have a flourishing career ahead of you.
The semaphore SMS system promises to make us healthier than alternative text input systems given the energy expended in sending just a few characters.
It occupies a cultural territory between Tai Chi, Cheerleading and Telecommunications which I don't think has ever been explored before.
The software saves a snapshot of each successful symbol match, including a picture of you and your message so far.
The composite video in step 1 shows a series of these images put together into a video and speeded up, which I did using an app called MPEGStreamClip. After each semaphore session, you'll find these images written into the program's directory.
Always happy to hear from people who want to take away these ideas and twist them to their own nefarious purposes. Look forward to hearing from you, and don't forget to rate this instructable if you like it.
Once the application starts up (or if the background seen by the camera changes significantly) it will take a few seconds to calibrate. During this time it will be confused and flash yellow and blue squares.
Once this has finished, you can move your yellow and blue 'flags' and the yellow and blue squares should faithfully follow your position. With the squares in the right position, just hold still and after about half a second, it will type the letter for you. If you need to repeat the same letter, this takes around two seconds.
If you make a mistake, signal 'DELETE" and it will remove the last character. When you've finished one stage of the message (e.g. dialling, writing the 'from', writing the 'subject' you can signal 'NEXT' to move on to the next stage. After you've filled out the body of the message, signalling 'NEXT" will send it, and then return you to the beginning again.
Semaphore is still taught and used in the Royal Canadian Navy, so you may even have a flourishing career ahead of you.
The semaphore SMS system promises to make us healthier than alternative text input systems given the energy expended in sending just a few characters.
It occupies a cultural territory between Tai Chi, Cheerleading and Telecommunications which I don't think has ever been explored before.
The software saves a snapshot of each successful symbol match, including a picture of you and your message so far.
The composite video in step 1 shows a series of these images put together into a video and speeded up, which I did using an app called MPEGStreamClip. After each semaphore session, you'll find these images written into the program's directory.
Always happy to hear from people who want to take away these ideas and twist them to their own nefarious purposes. Look forward to hearing from you, and don't forget to rate this instructable if you like it.
| « Previous Step | Download PDFView All Steps | Next Step » |
4
comments
|
Add Comment
|
1
Feb 22, 2008. 2:04 PMmasterme120
says:
I had an idea that might be possible with processing. Since it can generate keystrokes, do you think that it can generate mouse movements? You could set up the camera in front of your monitor, then have a stylus like object with leds in the end and 2 tactile switches, so that you could turn your computer into a tablet? When you push each tactile switch, the leds flash in a certain order, so that processing can detect from a hyper cam where you pushed the button? That may be way to complicated, but it's just an idea.
Reply
1
Dec 10, 2007. 11:49 AMmasterme120
says:
How would you make it use iChat?
Reply
3
Dec 10, 2007. 4:51 PMcefn (author)
says:
The way to connect the two is using the java.awt.Robot class, which is able to generate native keystroke events. So if you begin the processing application, then switch to iChat, the keystrokes get sent to iChat instead. Great idea, masterme!
Reply
3
Dec 10, 2007. 4:50 PMcefn (author)
says:
(removed by author or community request)
1
Dec 10, 2007. 5:15 PMmasterme120
says:
Could I try the prototype?
Reply
3
Dec 19, 2007. 9:36 PMcefn (author)
says:
Did you get a chance to test the keyboard prototype?
Reply
3
Dec 10, 2007. 6:06 PMcefn (author)
says:
I couldn't help it. Had to spend a bit of time making it a bit more solid and this is probably pretty well finished now. See the link in step one of the instructable. To send 'Enter' you have to hit 'Next' twice. Hitting Next just once switches you between numeric and alphabetical. The procedure is to launch the app in processing, then switch to iChat. All keystrokes sent are then captured by iChat instead.
Reply
1
Feb 22, 2008. 1:57 PMmasterme120
says:
I just tried it! It works perfectly, but my camera is not very good, could you please add manual brightness and contrast adjustors?
Reply
 |
Add Comment
|
