Re: 360 degree video real time... interested?

[Dave <dju@xxxxxxxxxxxxxx>]

On Fri, 29 Apr 2005 11:28:10 -0400, "z"
<rmwbsee@xxxxxxxxxxx> wrote:

>Thanks Dave
>That got me worried again,
>What is your def of patent squatting,
>How am I limited/affected by this method.

Hi Ryan,

I'll try to explain what I mean by patent squatting...in a
long roundabout way...<g>

I didn't mean to cast a pessimistic shadow over the
invention. I just wanted to express my views on what the
performance was that I witnessed. The camera I worked with
was a self-contained type with the camera unit (small CCD
type with focus lens) inside the housing.

Unfortunately, the CEO that I worked for didn't really have
a feel for R&D, nor sales, IMO. I tried to get her to
consider making something that the company could sell and
she just wanted to use her camera to capture some funding
from whoever might want to pay her. I was her only
programmer and the thing I wanted to make was something that
could be sold. IMO She wanted to sell vaporware (by asking
for funding without having a plan for how to return the
value to the fund  provider). My biggest gripe about that
was that I knew that software takes quite a while to develop
unless you've already written something similar
before....and I wanted to get started working on the
software since I knew it would take a while. She wanted to
either get someone at another company to provide funding to
develop whatever using the camera as the heart of the
project, or sue someone for patent infringement if the
opportunity comes up. But she wasn't interested in getting
started on a serious piece of software that accomplishes
something for an end user. So I ended up quitting because
there was a trust issue with her, plus I had trouble
collecting my pay and other employees had the same problem.

I tried to convince her (not being much of a salesman
myself) to sell kits that included software and hardware to
make it easy for an experimenter or manufacturer to make use
of the camera technology. She had no idea what to make
except that whatever the software does, it needs to 1:
support wireless transmission of video, 2: provide a way to
see a pano instead of the raw polar frames. That was the
limit of her ideas. A CEO without a vision, except to get
funded by someone, somehow. So I wrote a demo for her
(without wireless support) and she showed it to some people
she brought in. The people were disappointed in the way it
looked. I was disappointed too because I couldn't figure out
any way for software to solve the clarity issues. It was a
limit in the video resolution that really ended up being the
most frustrating problem. Going to a progressive scan
expensive type of camera might be one answer.

>
>My prototype [thread 3] is motion capture [timestamp is slow ftp issue]
>using non optic surfaces @ x10 resolution , it does improve with cam corder
>at higher res. giving 3-4 fps w 1.6ghz cpu, ok for me considering intended
>applications and that it was VB6. I am bsee not bscs. I'll have it optimized
>later.
>

I was trying to find a way to improve the conversion speed
and I came up with a couple of ideas at the time.

1: Convert part of the polar to rectangular but not the
entire polar image.

2: Try to find a way to replace the floating point trig Sine
and Cosine functions with fixed point. I tried to solve this
one and failed to improve the perf after writing a fixed
point tabel-lookup sine and cosine. Problem with that was
that the time it took to do a table lookup of the answer was
so long that it ended up taking longer than using the
floating point in the first place. That's the way R&D goes.

I'd be glad to talk more about it offline if you want. Feel
free to email me at yah oo at dav ee dot com. (remove the
blanks and change 'dot' to '.').

Regards,
Dave