Author Topic: Dan Danforth Molecular Dissociation of Water Document Inversion  (Read 864 times)

0 Members and 1 Guest are viewing this topic.

Offline Login to see usernames

  • Moderator
  • Hero member
  • ****
  • Posts: 789
I recently posted a diagram of the Danforth Project p. 19 and inverted the image to improve readabilty
of the part values or descriptions but there are apparently other techniques. Based upon the number
of pages in preceding articles and allowing 2 pages in the Space Energy Association magazine
for fly leaf and table of contents the page 19 comes from the magazine's page number , not
some other article or document that has 19 pages. I do not have the original  issue yet to verify this

"Standard Methods

"For data recovery from a document of low resolution, an inversion from white to black can often make
text or drawings more readable. One method that works well is to invert the document so that the
background is black and the characters or images are white. Sometimes this is all that is needed.
For more stubborn problems, shifting the white to a yellow green (based one the eyes greater
sensitivity in this color range) may gain the required information.

For example, if one looks at the moon at night, much detail is seen. The same moon with the same
amount of solar illumination when viewed at dust or early morn, much detail is lost.
The background sky floods the eye. Thus a white image on black background is often used in powerpoint or
slide presentations."

( Many of the 74 slides used in Stan's Pentagon presentation were white on black.) ed.

"Also, in darkened rooms the pupils dilate allowing for better visiual acuity."

(In the 1993 ISNE Colorado Symposium, the overhead lighting and generally greater amount of illumination
 in the meeting room washed out much of th imagery on many of the slides.) ed.

"Converting a document by inversion, applying edge enhancement filters, runniing an OCR program
such as Omnipage( to identify questionalbe characters) and then overlaying with a yellow green color substituion
for the white layer, mask out large area of brightness and then view in a room when your eyes have accomodated to the dark.
Have several members of the document recovery team view the document and compare the varied interpretations of images or text."
"There are cases that arise where the resolution of the image is so poor that a single pixel may be as large as text character.
But there is a method that can still obtain usefull information. Imagine,  if you will ,a rugby or basketball scoreboard ( a light matrix of of bulbs 5 by 6 for the numerical display.
The number 1 consists 6 lights lit in a vertical row, a number 7 might have 10 lights lit, the number 8 might have 21 lights lit, a space
with 0 lights lit. By examining the light intensity, certain numbers may be identified or eliminated from the likely underlying clear text.

"With text, the procedure is similar, but the use of gamma dictiionaries eg.  light, .medium dark, very dark,  very light
or intensity 7,18,14,0 might stand for the word ION,(space), With  the use of gamma dictionaries that are indexed to words and using clues from word length
and probable words  an underlying text is sometimes recovered. Pixel translation in the x axis or a straddle between two characters can also  be used
in diffiicult cases.. Additionally frequency analysis of the various pixels of similar shade may be fruitful.. Where the display maxtrix has fewer
elements such as in the nearly obsolete "nixie" tube, there are fewer gradations of light between various digits making analysis of covert images
of laboratory equipment displays more difficult".

"This process is similar to the use of CV pattern analysis for a simple monoaphabetic ciipher close to the unicity point. But in this case, probability
of the character is assigned on the position in a grey scale which can be calculated in advance by presumption of an assumed character
set and defining the matrix grid dimensions for the pixel. By changing the pixel height ,letters which are capitlal or letters which have portions
of the letter that fall below the base line of the text or above the mean height can sometimes be identified. For example an "h" , would have a
grey scale similar to an "i" for the top third pixel, a grey scale similar to and "n" for the milddle pixel and a grey scay similar to a space for the bottom pixel"

"In the case of license plates where low level lighting is often the case,  frame overlay from multiple video frames may allow for sufficient
edge contrast for numbers to be identified."
Анализ на документи Раздел 4

I know of two other HHO experimenters who are  using using inversion or photoshop filtering to decipher schematics
The use of pantone color matching can be used to identify resistor values  and  manufacturers of variious component parts
such as capacitors and inductors  ;)
« Last Edit: December 27, 2012, 13:51:42 pm by jim miller »