Re: ReL camera that talks
- From: "W. Nick Dotson" <nickdotson@xxxxxxxxxxxxx>
- To: "optacon-l@xxxxxxxxxxxxx" <optacon-l@xxxxxxxxxxxxx>
- Date: Fri, 07 Jul 2006 09:37:51 -0500
Right, the "threshold" know is the determinant of at what point the binary
decision of whether or not a dot exists/is detected and should be displayed.
That
being said, there is considerably less of what we would characterize as the
type of processing an OCR engine is having performed on the image prior to
getting the image, and desegmenting the page Etc... In the KT/NFB device, the
preprocessing isn't done by the OCR engine at all, but rather by the
software in the device which desegments the page... What I was getting at, is
that on a largescale image, I would want the device to merely buffer the
image, and allow me to scroll around on it, perhaps only with today's
technology, if there were a denser array possible, being able to apply
different levels of
"intensity" on a pin by pin basis, to attempt to convey some of what sightlings
see as "shading" in the image, if a user prefered that to the traditional
binary,
the dot is or isn't there model... Yes, I know about the different ways one
can currently access the display including the not mentioned "Weasel", but I'd
prefer something that could access the machines visual display data, without
having to physically wield the lense module, I've earned Tunel Carpel from
years of those maneuvers, with arthritis, as I never had the luxury of the face
up monitor...
Nick
On Fri, 07 Jul 2006 10:18:42 -0400 (EDT), Charles Pond wrote:
Hi Nick. Answers:
1. An array with different/lesser tactile parameters was not the
point. The size of the array if one were built today is not what
I meant; it is the size of a new, stand-alone optacon which could
be like a cassette container. The array can be as large or as
small as one wishes. As for cost, the cost of a new optacon could
be less than the original price, but the tactile display would
still be expensive. Nowadays, with better materials to make the
bimorfhs, and dedicated driver chips for anything like this problem
(backplane drivers for visual displays, drivers for MEMs), the
display could be smaller and use less power than those in our
optacons, but the cost might likely be around a thousand or so Cdn
(wild guess).
2. Connecting a tactile array to an OCR thing would not be as big
a deal as one would think. Whether trivial or not is not a
distracting concern for creative designers.
Connecting a tactile array to an oCR system (my Braille display is
connected to my computer) has the benefit of allowing the consumer
to add on what they wish, when they wish and at a pace affordable
to them. I doubt that a stand-alone optacon would sell nowadays,
but a module might. Of course, the access device MUST have a what-
you-feel-is-what-you-get mode for real-time optacon-like use. I'd
probably use that mode more than other image processing or image
enhancement modes. At this point, let me deal with the false
belief on the list that says that somehow the optacn simply pipes
the exact camera image directly to the tactile array, photosensor
to tactile stimulator, albeit there is a one-to-one mapping. This
is the impression which we have. Although there is no image
OCR'ing as with a package like K1000, there is signal handling and
digital decisions being made by the optacon's circuits which create
and affect what we feel on our finger. What we feel is actually
rebuilt and scanned onto our finger. It is not a simple straight-
through feeding of a tactile image to the display. The camera is
actually scanned to see which photosensors are excited (excites me
too). Also, "decisions" are made by the comparator circuitry in
the threshold feature to decide how bright or dark parts of the
image are relative to each other, and notably to a preselected
electrical level, which is then displayed for us as either
vibrating or passive pins. We perceive, quite accurately, what the
camera "sees", but we little know the stuff which goes on in the
black box to make it so. Even the image which we feel on our
finger is scanned on" electronically speaking, but done so quickly
that we perceive it as a solid, spontaneously fluid process of a
tactile shape. If Richard or Dave or someone like this is on the
list, they can clarify or correct me here. So, the tactile image
which we feel is actually scanned at several stages, and rebuilt
for us to perceive. So, what is the technical objection to a
digital component to a new real-time optacon system?
3. What we (you in this example) want can be done with a real-time
optaconer, whether built in 1973 or 2008. But for the sake of
being the Adversary's Advocate (whose days are short as the Bible
says): if a picture of something is snapped, reading it later on
won't change the information on the original image. It won't have
changed by the time we get around to reading it a few seconds
later. Speaking in milliseconds delay time, the optacon itself
stores image pieces (well actually "bits") in various shift
registers and latches until it is ready to have the tactile display
reflect them. so, whether we're talking in minutes, seconds or
milliseconds, the issue is the same, philosophically.
As for reading a BIOs:
1. You `an use an optacon with a crt lens or with the light of the
standard one shut off.
2. You can do a screen dump to a printer and read the printout.
3. You can send it to a Braille notetaker (the F.s. ones are
be>ter for this than the Pulsedata ones I feel), and read the file
while counting your keystrokes. Yep, life is a challenge. :)
Charles
---
From: nickdotson@xxxxxxxxxxxxx ("W. Nick Dotson")
Subject: Re: Camera That Talks
1. The tactile array had a specific matrix, and granularity
because scientifically conducted medical studies proved the
efficacy of an array with the characteristics used in the Model R1
devices. Those of us who use it heavily, for the mostpart, were
considerably less impressed with the R2's array. What
if any benefit could be garnered from an array that was broader
than that of the 2-fingered Japanese model? And, what is going to
drive down the manufacturing cost of the array?
2. Linking an OCR-Oriented device, which does alot of
preprocessing before OCR'ing the image, to a tactile array is not
going to be a trivial task, and to
what end would such a linkage be beneficial?
3. I do like the idea of being able to explore a computer screen
text graphics and all, as one sort of could with the Model 2's,
but, I'd want something freed of the constraints of having to
necessarily be tied to specific OS'(s), if possible, perhaps even
something giving one access to the BIOS, but those are
almost mutually exclusive of hardware and software dependencies...
Nick
On Thu, 06 Jul 2006 16:48:12 -0400 (EDT), Charles Pond wrote:
rather than simply rebuilding the optacon using the original
circuit
designs with today's technology, it might be sensible to build a
USB-compatible tactile array in a box with the needed controls,
and plug
it into something like this "camera that talks". I other words,
alter and
refine oleg's readiog machine-based approach. Does it make sense
to
re-develop a stand-alone optacon? From what I am piecing together
over
time of how the optacon R1D and II work (here a little, there a
little),
and with today's technology, it would seem feasible to build an
optacon
about the size of a cassette tape case or thereabouts. (maybe not
including battery?)
Charles
Shown 74%, press <SPACE> for more, 'q' to quit, or 'h' for help
To unsubscribe at any time, just send a message to:
optacon-l-request@xxxxxxxxxxxxx with the word "unsubscribe"
(without the quotes) in the message subject.
Tell your friends about the list. They can subscribe by sending
a message to:
optacon-l-request@xxxxxxxxxxxxx with the word "subscribe" (without
the quotes) in the message subject.
To unsubscribe at any time, just send a message to:
optacon-l-request@xxxxxxxxxxxxx with the word "unsubscribe"
(without the quotes) in the message subject.
Shown 93%, press <SPACE> for more, 'q' to quit, or 'h' for help
Tell your friends about the list. They can subscribe by sending a
message to:
optacon-l-request@xxxxxxxxxxxxx with the word "subscribe" (without
the quotes) in the message subject.
End of File, Press RETURN to quit
To unsubscribe at any time, just send a message to:
optacon-l-request@xxxxxxxxxxxxx with the word "unsubscribe" (without the
quotes) in the message subject.
Tell your friends about the list. They can subscribe by sending a message to:
optacon-l-request@xxxxxxxxxxxxx with the word "subscribe" (without the quotes)
in the message subject.
To unsubscribe at any time, just send a message to:
optacon-l-request@xxxxxxxxxxxxx with the word "unsubscribe" (without the
quotes) in the message subject.
Tell your friends about the list. They can subscribe by sending a message to:
optacon-l-request@xxxxxxxxxxxxx with the word "subscribe" (without the quotes)
in the message subject.
Other related posts:
