 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to the compensation of a video image to
provide an accurate black-and-white reproduction of the tonal luminance
differences of the image.
2. Description of the Prior Art
At the present time, in a number of fields, a photograph is made from a
video image, or the video image is part of the image reproduction chain,
and it is desired that the photograph be an accurate reproduction of the
tonal luminance differences of the image. For example, in the medical
field a CAT X-ray scan, an ultrasonic scan, an NMR scan (nuclear magnetic
resonance) or thermograph image is produced on a video monitor CRT
(cathode ray tube) screen for immediate viewing by the physician.
Simultaneously the same image is produced on another video CRT screen
which is part of an electronic camera photographic system. That electronic
camera photographic system includes the CRT tube, the video electronic
system to produce the image on the CRT screen, optical lenses and
equipment to focus the image and allow accurate exposure, and a
photographic sensitive material to take a photograph of the image which is
on the CRT screen.
The photograph is taken on photosensitive film that is developed using
conventional black-and-white (or color) film development processes. The
developed film may be directly viewed by the physician using a light box
which illuminates the film. Such light boxes are often non-uniform in
their illumination across the film and differ greatly in illumination from
one box to another.
In the printing field it is sometimes desired to make an accurate printed
picture from the image on a video monitor screen. For example, the screen
may show a computer-generated image. That image is reproduced on a
photosensitive material or directly on a printing plate to print a hard
copy that should be similar in tone, luminance and color to the original
object or scene.
It has been found that a photographic image reproduction may vary
significantly from the original image on the video monitor screen. Some of
the distortions are due to the inaccuracy of the CRT screen and video
image reproduction process. That type of distortion has been recognized
and compensation methods have been suggested, generally dealing with the
problem as it affects an entire group of video monitors. Similarly, the
distortions due to the camera have been recognized and treated, generally
by improvements in the optics of the camera or overall corrections in
video components (brightness, contrast, etc.). However, many of the
distortions are not predictable and may vary from day to day and from one
device to another.
Set forth below is a discussion of the problems most frequently encountered
in producing an accurate picture in the video-to-photographic process as
it relates to accurate tonal black-white reproduction. The contribution of
each problem to the total final distortion of the picture can change in
its characteristics periodically and is not predictable. One-day film
development bath temperature may be incorrect and seriously distort the
picture, and the next day it may still be incorrect but have only a minor
adverse effect due to partial compensation distortions from other
components in the system. In addition, most of the problem-causing effects
are non-linear, so that complete compensating for them in a simple direct
way is impossible.
The problems with the conventional system are explained in connection with
FIG. 1, which is a block diagram of a conventional black-white
photographic system. As shown in FIG. 1, the video image is produced by
the video source 10, which may be a video camera, a computer graphics
output, or a VCR. The video signal is viewed directly on the monitor CRT
screen 11. The same video image is shown on an internal CRT screen 13 in
the electronic camera 14. Generally the image on screen 13 is a negative
image compared to the image on monitor screen 11. The camera 14 includes
an optical system to take a still black-white or color photograph on the
film 15 which is removed from the camera 14, after the series of
photographs is taken and developed in a film processor 16. The film may
itself be the final hard copy 17 or may be used to produce a black-white
print using conventional print processing methods.
Each step of this conventional process gives rise to unpredictable
distortions. The first set of distortions arises in the CRT device, and
its screen 13, which is part of the electronic camera 14. The ratio
between luminance values, i.e., the ratio between shades of gray, on the
screen 13 may be inaccurate. For example, the CRT tube may be unevenly
coated with phosphor, or may be aged or may be subject to flare. In
addition, the relationship of the signal voltage applied to produce a
certain brightness is not linear. Consequently, the negative image
produced on the screen of the electronic camera may not be directly
proportional, i.e., accurate, compared to the positive image on the
monitor screen. A detailed description of the inaccuracy of an electronic
camera due to CRT distortions is found in Schwenker, R. P., "Film
Selection Considerations For Computed Tomography and Ultrasound Video
Photography": Proc. SPIE--Appl. of Optical Instrumentation In Medicine,
VII, 1979; 173, pgs. 75-80.
The electronic camera takes a picture using conventional black-white
photo-sensitive film. The film density, in such film, does not accurately
reproduce the differences in the gray scale because the film has a
non-linear "characteristic curve" of density against log exposure. The
exact shape of the curve varies from one manufacturer to another and even
from one batch of film to another, see The Theory of Photo Process, T. H.
James, pgs. 501-505, 4th Edition, Macmillan.
When the film is developed, distortions may arise from the variability of
the process chemistry, variations in process temperature, variations in
the film, and the non-linear characteristic of the photosensitive material
of the film. In those cases in which the film is duplicated or made into a
print, additional distortions may occur.
If the developed film is made into a print using a printer, still other
distortions may arise from the dot size of the printing, the spread of the
dots and the variable absorption of the ink into the paper due to various
types and batches of paper. Also, the perception of the gray scale may
differ depending on the type of printing process that is used.
After the film or other hard copy is produced, it is viewed under
conditions which may detract from the accuracy of the gray scale tones.
For example, the film may be placed on a light box whose intensity of
illumination is greater at its center than at its sides. Another cause of
viewing distortion is the "flare factor" in which flare (non-image light
from outside the image) enters the optical viewing system and mainly
affects the shadow areas.
The present invention is particularly directed to accurate reproduction of
the luminance differences in value (differences in a gray scale) and
absolute luminance on a black-white video screen. However, in its broader
aspects, the invention is also applicable to the accurate reproduction of
color images. The invention is directly applicable to color images in the
sense that the video screen may be a color CRT screen and the invention
will correct for gray scale distortions in reproducing the image on the
color screen. In addition, the reproduction of color images has its own
set of problems and distortions, aside from black-and-white tonal
differences. These color distortions can also be corrected, and their
correction will be discussed at the end of the detailed description.
These color and luminance distortions include (i) that the original color
is not exactly matched to the phosphors on the CRT screen so that the
color on the screen does not match the original color, (ii) that the color
of the photo-sensitive dyes of the film do not match the color on the CRT
screen and do not compensate for the color mis-match of the screen
phosphors, (iii) that the color of the color photo-sensitive papers, dyes
or printing inks do not match the color of the film. In addition, the
chemistry for color films and color prints is more complex, and more
temperature sensitive, than for black-white film and prints, so that
variations in the chemistry or temperature cause distortion shifts in the
color. A further problem with color, not found in black-white images, is
that the perception of color of the object or video screen (by the human
eye differs from the actual color on the film or print.
In U.S. Pat. No. 4,263,001 entitled "Apparatus and Method For Enhancement
of Optical Images", in one embodiment, which is not claimed, a video
camera is connected to an electronic image modification device which, in
turn, is connected to a single frame storage, to prevent feedback, and a
monitor CRT.
In U.S. Pat. Nos. 4,492,987 and 4,520,403, both entitled "Processor For
Enhancing Video Signals For Photographic Reproduction", the screen of an
electronic camera is electronically modified to enhance photographic
reproduction. The entire screen is treated as a unit and its brightness or
color is changed in accordance with the distortion introduced by a
selected photographic film.
In U.S. Pat. No. 4,658,286 entitled "Method and Apparatus For Correcting
Distortions In Reproducing Systems", a type of feedback system is
described. In one embodiment three photocells look at a corner of the CRT
screen having test colors and their outputs are compared to reference
colors.
OBJECTIVES AND FEATURES OF THE INVENTION
It is an objective of the present invention to provide a more accurate
black-white photographic image taken from the image on a video screen in
which the photographic image more accurately maintains the relative and
absolute (for luminance reproduction) tonal scale of gray tones.
It is a further objective of the present invention that the video image be
compensated to obtain the accurate photographic image and that such
compensation takes account of short term, for example, daily distortions,
and long-term distortions and fixed sources of distortions.
It is a still further objective of the present invention that the operator
is alerted to excessive deviations from a standard of tonal reproduction
so that he may take immediate corrective action to restore the system so
that it will produce reproductions with accurate tones.
It is a feature of the present invention to provide a method and system for
the accurate tone reproduction of the luminance ratios in a black-white
image. In one embodiment the original image may be viewed on a monitor CRT
and it is also shown on the CRT screen of a video device which is part of
an electronic camera. The video device has means to vary the luminance
values (brightness-darkness) of each pixel on the camera CRT screen. A
test image is shown on that CRT screen, the test image having areas
differing in luminance, for example, a ten-segment gray scale. The test
image is photographed by the camera using the same batch of film as will
thereafter be used and the film is developed using the same chemicals and
conditions as will thereafter be used. A photoelectric densitometer is
used to test the densities of the test pattern on the developed image. The
density values, in digital format, are entered into a computer having a
look-up table whose entries are compared to table entries representing the
ideal luminance for each tone. The look-up table provides a compensation
value for each of the original camera or CRT tones.
When a certain pixel is to be activated on the CRT screen, its original
brightness value is compensated for by the compensation value.
Consequently, for example, an original lighter gray tone may be
compensated to become brighter and an original darker gray tone may be
compensated to become darker. In this way, in the same video frame, some
pixels are made brighter and some pixels are made darker, in order to
compensate for the distortions arising from the reproduction process.
SUMMARY OF THE INVENTION
The present invention uses the same components as a conventional system
and, in addition, uses additional means to provide a rapid compensation so
that the black-white tones (luminance differences on a video screen) are
accurately reproduced.
The system, in one embodiment, uses a video source, such as a computer
graphics output, VCR or video camera to produce an image on a CRT monitor.
In one embodiment, the video source simultaneously also produces a
reversed image on a CRT screen of an electronic camera. The electronic
camera takes a photo-sensitive film picture of its CRT screen and the film
is developed using conventional film processing.
The CRT video system of the electronic camera is connected to a computer
having a look-up table memory and an image to be seen on the screen. The
computer's image memory processes each frame of the video image on a
pixel-by-pixel basis in digital form. A black-white gray scale test
pattern is shown on the CRT screen of the electronic camera, having
preferably at least 10 gray-scale areas. The test pattern is reproduced as
a latent image on film in the electronic camera and developed into a
negative image as hard copy. The test pattern is then sensed by a
densitometer which provides a electrical signal corresponding to the
gray-scale density of the test pattern on the developed film. The
densitometer's output, in digital form, is entered into the computer. The
computer, using its look-up table memory, will determine the required
compensation, on a pixel-by-pixel basis. That compensation is applied to
each video frame which passes through the computer's image memory.
The computer memory includes an ideal set of density value corresponding to
the luminance values for each tone. Those ideal values are compared to the
actual values, from the densitometer, to provide the required
compensation. The compensation is non-linear so that making the entire
video screen darker or brighter will not compensate for the distortions in
the ratios between the tones. Instead, each tone requires its own
compensation. Since an image is composed of different tones in different
areas of the video screen, the luminance (brightness) value of each tiny
area (pixel) is individually compensated to produce the tonal reproduction
accurately.
Other embodiments of the present invention include other types of hard copy
reproduction systems in place of an electronic camera. Such alternative
systems include: (i) an xerography system in which a latent image
corresponding to an image on a CRT video monitor is formed by a laser beam
on a photo-sensitive drum; (ii) a laser-film system in which such a
corresponding latent image is formed by a laser beam directly on
photo-sensitive film; (iii) an ink-jet system in which the corresponding
image, in this case a visible image, is formed by ink dots from an ink jet
printer on paper; and (iv) a thermal printer system in which dyes or other
imaging materials are transferred to a substrate, or activated in a
substrate, to show a visible image, after processing.
BRIEF DESCRIPTION OF THE DRAWINGS
Other objectives and features of the present invention will be apparent
from the following detailed description of the inventor's presently known
best mode of practicing the invention, taken in conjunction with the
accompanying drawings.
In the drawings:
FIG. 1 is a block diagram of a prior art system to produce photographs of a
video screen image;
FIG. 2 is an X-Y graph showing the film density in the reproduced image
plotted against camera video screen log luminance, i.e., different pixel
values;
FIG. 3 is a block diagram of one embodiment of the system of the present
invention;
FIG. 4 is an X-Y graph in which the reproduction image (luminance) (Y axis)
is plotted against screen luminance (X axis);
FIG. 5 is an X-Y graph in which log original screen image luminance on the
Y axis is plotted against the screen pixel values (0-256) on the X axis;
FIG. 6 is an X-Y graph in which the reproduced image density on the Y axis
is plotted against the log original monitor video screen intensity
(luminance) and screen pixel values;
FIG. 7 is an X-Y graph in which the reproduced image density on the Y axis
is plotted against the original screen pixel values on the X axis.
DETAILED DESCRIPTION OF THE INVENTION
The present invention relates, in the below-described first embodiment, to
the accurate reproduction of the black-white tone from the CRT screen of
an electronic camera. The "tone" of an image is the ratio between its
luminance values or brightness of the screen image. That tone ratio is
sometimes called "luminance differences" or "shades of gray" or a "gray
scale".
An accurate reproduction of the ratio of luminance values (gray scale)
would occur when the ratio of density units on the film directly
corresponds (linear relationship) to the gray scale on the screen. The
density units on the film is a measure of the blackness of the film.
Density is defined as the negative log transmission of the light which
passes through the film or log reflectance of an opaque substrate. The
"screen luminance" is the brightness of the CRT screen, or a portion of
the screen, and is measured in terms of "log screen luminance" to directly
correspond to density which is also a log function.
The CRT screen is divided into "pixels" with the pixels arranged in columns
and rows. The number of pixels which is selected depends on the original
image generator, i.e., medical scanner output, the screen size, the
electronics of the CRT device (video electronics) and the video system
which is used. In the United States there are various standards of video
systems for medical purposes. For convenience, we may consider a system of
1000 horizontal lines and each line may be considered as 1000 pixels so
the total number of pixels comprising the CRT screen is 1000.times.1000,
or one million pixels. As explained below, using digital image processing
device (memory store) each pixel value, it is possible to change the
back-white value of each pixel. In the case of an 8-bit digital system,
each pixel has a black-white gray scale of 256 values from 0 (blackest) to
255 (brightest), on the monitor screen.
As shown in FIG. 2, it is relatively simple for the reproduced image to
have the required minimum density (D min) and maximum density (D max).
However, the gray scale on the film image is a non-linear curve C which
differs from the ideal gray scale 1 which is linear. FIG. 2 plots the film
density in the reproduced image against the CRT screen of the electronic
camera in log luminance (brightness).
The present invention approaches the ideal gray scale of curve 1 by
electronically adjusting the value of each pixel. If the film density is
below the ideal curve, then the luminance of the pixel is raised, i.e.,
the brightness of the pixel is increased, so that the final density of the
dot on the film corresponding to the corrected pixel on the electronic
camera screen, after such compensation, is on the ideal gray scale line 1.
As shown in FIG. 3, a block diagram of the first embodiment, a monitor
video 20 includes a CRT screen. The monitor 20 shows an image which in
this embodiment is a black-white image. Alternatively, the image may be a
color image in which case correction of the tone on the reproduction will
by itself, and without correction of color, greatly improve the accuracy
of the color reproduction, for example, an 80% improvement.
The monitor 20 is preferably a high resolution video monitor having 1000
horizontal lines per video frame, each frame consisting of two interlaced
fields each of 500 horizontal lines. There is no attempt, in this example,
to enhance the image on the monitor 20; but only to enhance the reproduced
hard-copy image so that it accurately matches the image on the monitor
screen. However, the same method and system may be used to improve the
image on another video screen (soft copy) so that it accurately reproduces
the image on the monitor screen.
The same image which appears on the monitor screen also appears on the
screen 22 of the video device 23, which is a part of the electronic camera
24. The electronic camera includes the optical system to focus the image
from CRT screen 22 onto the photosensitive film 25 in its camera body 26.
In a photographic negative system the film 25 is preferably a high
resolution black-white negative film. The image on CRT screen 22 is
preferably the inverse (inverted black-white) of the image on the screen
of the monitor 20.
The electronic camera 24 is connected to the electronic processor 30 which
is the computer means to calculate the correction values and to control
the luminance of the CRT screen on a pixel-by-pixel basis. In this
example, the CRT screen with its 500 horizontal lines per field (500 per
frame and 2 fields each 1/60th second) is preferably divided so that each
line has 500 pixels for a total of one million pixels per frame. Video
images are conventionally generated in an interlaced fashion wherein every
other line in the image frame is generated during one field interval and
the alternative lines are generated during the next field interval.
The processor 30 includes a digital microcomputer, for example, intel,
having a CPU (central processing unit) 31. The processor has three digital
memories, which are preferably solid-state VLSI chips (Very Large Scale
Integrated). The first memory 32 is a read-write LUT look-up table which
includes, as one set of data held therein, the correct predetermined
brightness (luminance) value that a pixel should receive to correctly
portray each tone of the gray scale ("ideal value").
The second memory is a video access memory 33 which stores at least one
field of the digital memory. Video access memories are characterized by a
serial access port through which the video data can be input and output
independently of other memory signaling and timing requirements. The
present state of the art in these components is a 64k.times.4 device solid
state integrated random access memory circuit ("RAM chip") with an
internal 256.times.4 serial access buffer. Suitable devices are Hitachi
(HM53461 or HM53462), Fujitsu (MB81461), and Mitsubishi (M5M4C264) for
NTSC video signals.
The film 25 is developed, in conventional manner, in the film processor 35.
A densitometer 38, for example, "X-rite 301" model, having a digita output,
is connected to the video processor 30. The densitometer measures the
density values of the test pattern 37 on the developed film 36. The test
pattern is preferably a band of 11 side-by-side panels of different
predetermined tones of gray forming a gray scale. Alternatively, the gray
scale may be obtained from a series, for example, 11, film exposures made
in sequence. The density measured on developed film 36 appears in a
digital electrical signal at the output of densitometer 38.
The digital values representing the actual densities of the gray scale of
the test pattern are entered into the computer 30 which forms a new
look-up table 32. For each actual density value of a gray tone there is,
stored in the look-up table 32, an ideal value.
The ideal value of the density representing the luminance of the monitor
screen is obtained as follows: The luminance value of any pixel value, at
any particular screen and ambient lighting condition cimbination) is found
using a spot photometer 41, for example, the Minolta LS-100. A graph of
pixel value Vs log original screen luminance is constructed (FIG. 5). The
graph of FIG. 6 is then constructed based upon the actual luminance values
of FIG. 5. In FIG. 6 the dual values of log intensity of screen and the
actual pixel values of these screen luminances are marked simultaneously
on the X-axis.
The density values of the film, i.e., the reproduction image density, as
measured by a densitometer, are shown plotted on the Y-axis of FIG. 6.
There is a linear relationship, in FIG. 6, shown by the ideal curve I,
between density and log original screen luminance. The actual curve A,
which are the densities as measured by the densitometer on the test
pattern, is not linear.
FIG. 7 is derived from FIG. 6. The various curves of the Figures are
intended primarily as illustrations of the principles. In practice, the
procedure is to use digital software (computer programs and memory) to
establish the ideal pixel values for each gray scale tone.
The graph of FIG. 2 is the basis to construct the graph of FIG. 6 showing
pixel values Vs (Ideal and Actual) densities.
The ideal pixel value is found on FIG. 7 as described above. Every screen
pixel value is associated with a corresponding camera signal pixel value.
The values of all the screen pixel values and their corresponding camera
pixel values are entered in the look-up table (LUT) memory 32.
Using the LUT memory 32, each screen pixel value is changed to the
corresponding camera pixel value. This correction yields an ideal density
on the film which follows the ideal (or desired) tone reproduction curve.
An example is as follows:
1. Generate a gray scale test pattern of known pixel values of distinct
steps (e.g., the SMPTE RP-133 of 11 steps of pixel values, in the form of
side-by-side bands. The preferred pixel values are: 0, 25, 50, 75, 100,
125, 150, 175, 200, 225, 255.
2. Measure the screen luminance values on the screen of the video monitor
20 of these pixel values with the spot photometer. Then photograph this
test pattern (screen image) with a camera calibrated to give the desired
Dmin and Dmax of the reproduced image (i.e., a well calibrated
camera-processor system). Measure the film densities on the developed film
using the densitometer. Construct FIGS. 5, 6 and 7 in that order.
3. From FIG. 7 construct the LUT 32. The LUT 32 will feed their values to
the computer program which programs CPU 31 so that each screen pixel value
is changed to the appropriate camera pixel value. This yields the ideal
density for the ideal tone reproduction.
The following table can be used for illustration:
__________________________________________________________________________
Column:
6 7 8
.DELTA. D
Pv .DELTA. Pv
1 2 3 change
required
change
test
screen
luminance
4 5 in (ideal)
in
pattern
pixel
value resulting
ideal densi-
(pixel)
pixel
step #
value
cd/m sq.
densities
densities
ties
values
values
__________________________________________________________________________
Log
0 0 0.93-0.03
2.25 2.25 0.00
0 0
1 25 1.04-0.07
2.09 2.2 0.11
13 -12
2 50 1.95-0.29
1.9 1.93 0.03
45 -5
3 75 4.2- 0.62
1.65 1.59 -0.06
83 8
4 100 8.9- 0.95
1.4 1.27 -0.13
113 13
5 125 15.6-1.19
1.2 1.02 -0.18
141 16
6 150 24.9-1.4
0.93 0.82 -0.11
163 13
7 175 36.5-1.56
0.73 0.66 -0.07
183 8
8 200 51.3-1.71
0.5 0.51 0.01
198 -2
9 225 70.0-1.85
0.36 0.37 0.01
220 -5
10 255 93.0-1.97
0.25 0.25 0.00
255 0
Values obtained from:
test
test
photometer
densito-
FIG. 6 or
Col. 5
FIG. 2
Col. 7
pattern
pattern meter
equaton
(-) or (-)
1 Col. 4
equation
Minus
2 Col. 2
__________________________________________________________________________
The following mathematical tools help in reducing most of the labor
described above:
1. Instead of construction FIGS. 7 and 6, the ideal density is calculated
as follows (to a close approximation):
Equation 1:
Ideal density=log (actual lumnance values)*[Dmax-Dmin/log luminance
max.-log luminance min.]+Dmax-(Dmax-Dmin | | |
