Re: [SLUG] A Challenge....

From: chris clement (chris@ozline.net)
Date: Wed Dec 18 2002 - 22:21:57 EST


Also, on the software side, I was just thinking of opening a text editor
before the sluice run, and saving it at the end. Then process the file by
converting the keypresses to coordinates and applying some math rules.
----- Original Message -----
From: "chris clement" <chris@ozline.net>
To: <slug@nks.net>
Sent: Wednesday, December 18, 2002 10:04 PM
Subject: Re: [SLUG] A Challenge....

> You are right. I discovered that right after my post. I can mash 3 in each
> row at the same time, however. It might still be adequate if we only
expect
> a few fish shadows from time to time and not coming thru like 'dines in a
> tin.
> ----- Original Message -----
> From: "Andrew Wyatt" <awyatt@fewt.com>
> To: <slug@nks.net>
> Sent: Wednesday, December 18, 2002 9:17 PM
> Subject: Re: [SLUG] A Challenge....
>
>
> > On Wed, 2002-12-18 at 19:55, chris clement wrote:
> > > I really like the keyboard idea. Talk about a cheap interface! I would
> use
> > > two rows of keys with one hooked up to a side array of beam detectors
> and
> > > the other to an overhead array. Now we have a grid of, say, 144
> > > intersections. I think the input buffer will catch all the keystrokes.
> With
> > > this little "CAT scan" output, it should be fun to figure out how many
> fish
> > > are in the beams together!
> >
> > Most keyboard buffers are limited in that they won't register more than
> > 6 keys pressed at the same time. (There has been a lot of research done
> > on the subject in the MAME world.) Capturing a single keystroke is easy,
> > in perl you would use something like this:
> >
> >
> > system "stty -echo cbreak </dev/tty >/dev/tty 2>&1";
> > $cntr=0;
> > for (;;) {
> > eval {
> > local $SIG{ALRM}=sub{last;};
> > alarm 10;
> > $key=getc;
> > alarm 0;
> > ++$cntr;
> > print "Captured char code ".ord($key).", counter is now
> > ".$cntr."\n";
> > }
> > }
> > system "stty echo -cbreak </dev/tty >/dev/tty 2>&1";
> >
> > -Andrew
> >
> > > ----- Original Message -----
> > > From: "Andrew Wyatt" <awyatt@fewt.com>
> > > To: <slug@nks.net>
> > > Sent: Wednesday, December 18, 2002 6:04 PM
> > > Subject: Re: [SLUG] A Challenge....
> > >
> > >
> > > > On Wed, 2002-12-18 at 16:21, wchast@utilpart.com wrote:
> > > > > This is a little project I am going to try to take on because
> > > > > the cost of a device available commercially is off on another
> > > > > planet.
> > > > >
> > > > > My wife has a live bait business, part of that involves delivering
> > > > > bait fish to bait and tackle shops. She has to count the buggers,
> > > > > and the count always is off regardless of WHO counts them. I
started
> > > > > looking around for a fish counting device and found one that uses
> > > > > a optical scanner to count them as they go through a sluice. The
> optical
> > > > > part uses IR LED's to illuminate, and the detector or scanner uses
> > > > > what appears to be a column of detectors on the other side of the
> > > > > sluice, as a fish goes by it the detectors see the change in IR
flux
> > > > > and can detect when one fish is ahead of the tail of the first
fish
> > > > > thereby not opening all of the optical path.The fish will be
> anywhere
> > > > > from 3-6 inches in size.
> > > > >
> > > > > The manufacturer of this device wants more than $5K for it, I
> suppose
> > > > > that is ok for some government operation that has deep pockets but
> > > > > at $0.25/fish she will have to catch a lot of fish to pay for it,
> she
> > > > > also has to pay a truck used to deliver them and the guy and boat
> who
> > > > > actually goes out and catches them.
> > > > >
> > > > > The hardware does not look to be that great a deal. It is probably
> in
> > > > > the software where the cost is.
> > > > >
> > > > > My idea would be to feed the data from the sensors into the
parallel
> > > > > port and read it at a rate such that I get each fish as it goes
down
> > > > > the sluice.
> > > > >
> > > > > Now my question is this, do any of you have a idea of where I
should
> > > > > start with this one. The hardware is a no brainer, I can get IR
> sources
> > > > > and detector arrays, I can build the sluice and other material,
but
> I
> > > > > will need to come up with some software that can count the fish as
> the
> > > > > go in front of the detector. Has anyone done anything similar in
> terms
> > > > > of software?
> > > > >
> > > >
> > > > Why don't you replace an old keyboard key with the IR
source/detector
> > > > and then grab the scancode that gets sent? There has to be a method
> that
> > > > will not send a keystroke unless the IR is broken. If you can use a
> > > > keycode you can have the software in minutes.
> > > >
> > > >
> > > > --
> > > > Andrew Wyatt <awyatt@fewt.com>
> > > > FEWT Software - http://www.fewt.com
> > --
> > Andrew Wyatt <awyatt@fewt.com>
> > FEWT Software - http://www.fewt.com



This archive was generated by hypermail 2.1.3 : Fri Aug 01 2014 - 19:35:16 EDT