Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
.3;~ J
SCENE BAI~ANCE CAI~IBRATION O~ DIGITAI~ gCANNER
FIEI.D Q~ TH:~ INVENTION
The present invention relates in general to
digitized color imagery photofinishing systems and is
particularly directed to a mechanism for using a scene
balance mechanism to calibrate the operation of a hign
resolution digital opto-electronic scanner.
~0
BACKG~OUND OF_THE I~NTION
Because the dynamic range of photo-imagery
input/output devices employed in photofinishing systems
is considerably narrower than that of the image capture
medium (photographic film), not all of the information
contained on the film is reproducible. In order to
optimize the presentation of a reproduced (printed)
image to the human visual system, the device (opto-
electronic scanner) that inputs the film image to the
photoprocessing system should be calibrated such that
the principal subject matter of the image preferably
falls within the linear portion of the response
characteristic of the imaging device. In an analog
photofinishing system, calibration is driven not only
by the limited dynamic range of the components, but it
can be expected that the response range of the film
scanner may not match or even substantially overlap the
output transfer function of the reproduction medium
(print paper). Consequently, calibration can become a
difficult and time-consuming trial and error process.
In a digital imagery processing system, on
the other hand, where the input device (a digital opto-
electronic scanner~ and the output device (e.g. high
resolution thermal printer) employ spatial arrays of
pixels that interface (via D~C circuitry) with digital
-2- ~f~3~.s~
data signals, the fundamental mismatch problem does not
occur. However, with the introduction of very high
resolution film scanners (e.g. those having an imaging
pixel array of 2028 x 3072 pixels, the response of
which is resolved into sixteen bits per color per
pixel), the quantity of data produced per image is so
large that it must be reduced for storage in a
practical sized framestore. Namely, some of the scene
information in the digitized image must be discarded.
Thus calibration of the scanner remains a key aspect of
quality system performance.
SUMMAR~ OF THE INVENTION
In accordance with a first embodiment of the
present invention, the above-referenced calibration and
storage problem of conventional photofinishing systems
is obviated by means of a new and improved digital
imagery capture and storage mechanism which scans the
image twice, the first scan being used to gather data
to calibrate the scanner, and the second scan being
used to capture and store a high resolution image.
For this purpose, the scanner may be
controlled to carry out a low resolution mode, prescan
of the color photographic image of interest, thereby
obtaining a low spatial resolution digitized image.
(Because fewer pixels are scanned during the low
resolution scan the low resolution pixel values may be
encoded to a greater data precision.) This low
resolution digitized image is then analyzed by a scene
balance mechanism to detPrmine how the response
characteristic of the scanner's imaging pixel array
sees the image and encodes its spatial content. For
purposes of the present invention, by 'scene balance
mechanism' is meant an adjustment of image color
balance based upon the scene content and the
-3-
sensitometric characteristics (e.g. exposure, light
source etc.) of the image being reproduced. The output
of this analysis, which represents the color balance
content of the digitized image, is then used to adjust,
or calibrate, the sensitivity parameters of the
scan~er, so that~ during a subsequent high resolution
scan of the image, the essential subject matter of the
image will fall within the linear portion of the
response range of the scanner's imaging pixel array.
Adjustment of 'color balance' is defined as adjusting
the average red, green and blue image levels, so as to
ensure that an image will have the appropriate color
and neutral reproduction characteristics.
The high resolution digitized image is then
processed by the scene balance mechanism to map the
image data into a digitized image having a reduced
encoding resolution corresponding to that of an
attendant framestore. Although some of the information
in the captured scene is lost by the reduction in
encoding resolution, since the scanner has been
calibrated so as to optimize its sensitivity to the
color balance content of the image, the essential
information (i.e. that which is necessary to reproduce
a high quality image) is captured and stored.
Pursuant to a second embodiment of the
invention, rather than scan the image twice, once at
low resolution for purposes of calibration, and then at
high resolution for data capture, the image is scanned
only once, at high spatial resolution and high digital
resolution. The high spatial resolution image is then
converted into a high digital, low spatial resolution
image, which is processed to calibrate the scene
balance mapping function. Namel~, the processed data is
used to calibrate the mapping of the originally
~ J~5~
digitized image into a reduced digital resolution (e.g.
eight bits per pixel per color) framestore.
BRIEF DESCRIPTION OF TH~ DRAWI~
Figure 1 diagrammatically illustrates a
photographic color film photofinishing minilab with
which the scene balance-based digital imagery capture
and storage mechanism of the present invention may be
employed;
Figure 2 is an imagery processing flow
diagram of the scene balance based calibration and high
resolution capture mechanism of a first embodiment of
the present invention; and
~ Figure 3 shows the steps of an alternative
calibration mechanism in which only a single (high
resolution) scan is carried out.
DETAI~ ES~RIEIION
Before describing in detail the scene
balance-based digital imagery capture and storage
mechanism in accordance with the present invention, it
should be observed that the present invention resides
primarily in a novel structural combination of
conventional digital imagery processing circuits and
components and not in the particular detailed
configurations thereof. Accordingly, the structure,
control and arrangement of these conventional circuits
and components have been illustrated in the drawings by
readily understandable block diagrams which show only
those specific details that are pertinent to the
present invention, so as not to obscure the disclosure
with structural details which will be readily apparent
to those skilled in the art having the benefit of the
description herein. Thus, the block diagram
--5--
illustrations of the Figures do not necessarily
represent the mechanical structural arrangement of the
exemplary system, but are primarily intended to
illustrate the major structural components of the
system in a convenient functional grouping~, whereby the
present invention may be more readily understood.
Figure 1 diagrammatically illustrates a
photographic color film processing system (e.g.
photofinishing minilab) with which the present
invention may be employed and, for purposes of the
present description, such a system may be of the type
described in co-pending Patent application Serial
Number , filed _ , by S. Kristy, entitled ~ulti-
resolution Digital Imagery Photofinishing System",assigned to the assignee of the present application and
the disclosure of which is incorporated herein. It
should be observed, howe~er, that the system described
in the above-referenced co-pending Kristy application
is merely an example of one type of system in which the
invention may be used and is not to be considered
limitative of the invention. In general, the invention
may be incorporated in any digitized imagery processing
and reproduction system.
In accordance with the imagery data
processing system of the abo~e referenced co-pending
~risty application, each high resolution captured image
is preferably formatted and stored as a respective
image data file containing a low, or base, resolution
image bit map file and a plurality of higher resolution
residual images associated with respectively increasing
degrees of image resolution. By iteratively combining
these higher resolution residual images with the base
resolution image, successively increased resolution
images may be recovered from the base resolution image.
-6~
As an example, spatial data values representative of a
high resolution ~3072 x 2048) image scan of a 36mm-by-
24mm image frame of a 35mm film strip may be stored as
a respective image data file including a base
resolution image bit map file containing data values
associated with a spatial image array or matrix of 512
rows and 768 columns of pixels and an associated set of
residual image files to be stored on the disc. Within a
photofinishing workstation, the base resolution image
may be further su~-sampled to derive an even lower
resolution sub-array of image values (e.g. on the order
of 128 x 192 pixels) for use by the photofinishing
operator in the course of formatting and storing a
digitized image file.
Thus, in the digital image processing system
of Figure 1, color photographic images, such as a set
of twenty-four or thirty-six 36mm-by-24mm image ~rames
of a 35mm color film strip 10, are scanned by a high
resolution opto- electronic color film scanner 12, such
as a commercially available Eikonix Model 1435 scanner.
High resolution film scanner 12 outputs digitally
encoded data representative of the response of its
imaging sensor pixel array (e.g. a 3072 x 2048 pixel
matrix) onto which a respective photographic image
frame of film strip 10 has been projected by an input
imaging lens system. This digitally encoded data, or
'digitized' image, is encoded to some prescribed
resolution (e.g. sixteen bits per color per pixel) that
~0 encompasses a range of values over which the contents
of the scene on the color film may vary. For a typical
color photographic negative, the range of values is
less than the density vs. exposure latitude of the
film, but is sufficiently wide to encompass those
density values that can be expected to be encountered
for a particular scene.
3"~
--7--
As noted earlier, because of its very large
(2048 x3072) spatial resolution, with the output of
each pixel being resolved to sixteen bits, the quantity
of data per image produced by such high resolution film
scan~ers is so large that it must be reduced for
storage and reasonably fast access in a practical sized
framestore, which necessarily implies that some of the
scene information in the digitized image will be
discarded. For this purpose, a scene balancing
mechanism is used to map the digitized image into a set
of lower resolution digital codes (e.g. eight bits per
color per pi~el), each of which has a resolution
corresponding to the dynamic range of a digitized image
data base (framestore). The database may be resident a
in photofinishing workstation 14, which contains
imagery application software through which the
digitized image may be processed to achieve a desired
base image appearance and configuration in the course
2~ of driving a high resolution thermal printer 16 to
output a high ~uality color print.
Preferably, in the course of being mapped
into memory, the digitized imagery data output by the
~5 high resolution film scanner is su~jected to a code
conversion mechanism of the type described in co-
pending application Serial No. , filed , by T.
Madden et al, entitled "Extending Dynamic Range of
Stored Image Database, n assigned to the assignee of the
present application and the disclosure of which is
herein incorporated. Pursuant to this code conversion
scheme, the dynamic range of the digitized image
database may be extended to permit shifting of encoded
pixel values without 'clipping', and to provide a
limited window of values into which extremely high
reflectance image points may be encoded and stored. To
~B- o~ J~
this end, digital codes, into which the high resolution
imagery data output by the image scanner are mapped by
the scene balance mechanism, are converted into a set
of reduced-range digital codes of the same resolution
5 as, but having a smaller range of image content values
than the dynamic range of the digitized image data
base. The code conversion mechanism operates to convert
a maximum value of 100% white reflectance to an encoded
value that is less than the upper limit of the dynamic
range of the database to accommodate shifts in the
digitized imagery data and allow for the placement of
specular highlights that are beyond the 100% white
reflectance maximum.
As pointed out above, when digitizing an
image on the film strip, it is preferred that the film
scanner be calibrated such that the principal subject
matter of the image falls within the linear portion of
the response range of the scanner's imaging pixel
array. For this purpose, a first embodiment of the
present invention employs a calibration and high
resolution capture procedure, diagrammatically
illustrated in the imagery processing flow diagram of
Figure 2, whereby the image is scanned twice, once at
low resolution for purposes of calibration, and then at
high resolution, for data capture.
More particularly, as shown at step 101,
image scanner 12 is controlled to carry out a low
resolution mode, prescan of an image 10 of interest.
Where the scanner has multiple resolution scan
capability, it is controlled so as to scan the image at
a relatively low spatial resolution, e.g. on the order
of seven to twenty-four by ten to thirty-six pixels per
frame. Depending upon the size o~ the low resolution
image, it may be necessary to perform a further spatial
_g_
compression of the captured image, in order to reduce
the computational intensity (and thereby achieve a
reasonably rapid throughput) of the application of the
low resolution image to a scene balance mechanism. In
accordance with the multiple mode operation of the
above-referenced high resolution scanner, during low
resolution scan, a 128 x 192 image is captured. Through
further spatial integration of the imagery data within
workstation 14, a captured 128 x 192 pixel version of
the image may be reduced to a very small sub-array
(e.g. 24 x 36 pixels, each encoded at sixteen bits per
color) for application to the scene balance mechanism
through which high resolution imagery data is mapped
into the framestore.
This very low resolution (24 x 36) digitized
image is then analyzed in step 102 by the scene balance
mechanism to determine how the response characteristic
of the scanner's imaging pixel array sees the image and
encodes its spatial content. The scene balance
mechanism (the image processing result of which may be
implemented as a set of look-up tables (LUTs ), one for
each RGB color) outputs three values, one for each
color, which represent the color balance content of the
digitized image.
In step 103, using these values, the
sensitivity of the scanner is calibrated, so that,
during a subsequent high resolution scan o the image,
the essential subject matter of the image will fall
within the linear portion of the response range of the
scanner's imaging pixel array. While the scene balance
output values may be employed to effect vernier
adjustments of reference voltages for the scanner's
imaging array, in accordance with a preferred mode of
the present invention, a respective offset code, one
10-
for each of the color values, is added to the inputs of
each scene balance look-up table in order to
effectively shift or translate its mapping function
that brings the essential subject matter of the image
into the linear portion of the response range of the
scan~er's imaging pixel array.
With the scanner now calibrated, (e.g. scene
balance look-up tables shifted to optimize the use of
the imaging array's linear response range), the scanner
is controlled in step 104 to execute a high spatial
resolution scan of the image. Since the scene balance
LUTs have been tran~lated in accordance with the output
of the low resolution prescan, the high resolution
digitized image will be mapped into the framestore such
that essential image information (i.e. that which is
necessary to obtain a high quality print) is captured
and stored.
Figure 3 shows the steps of an alternative
calibration mechanism in which only a single (high
resolution) scan is carried out. Pursuant to this
embodiment, rather than scan the image twice, once at
low resolution for purposes of calibration, and then at
high resolution for data capture, the image is scanned
only once, at high spatial resolution and high digital
resolution. The high spatial resolution image is then
converted into a high digital, low spatial resolution
image, which is processed to calibrate the scene
balance mapping function. Namely, the processed data is
used to calibrate the mapping of the originally
digitized image into a reduced digital resolution (e.g.
eight bits per pixel per color) framestore.
More particularly, with reference to Figure
3, at step 200, the image is scanned to obtain a high
3 ~-4i,~
spatial resolution (e.g. 2048 x 3072) image digitized,
for e~ample at sixteen bits per color per pixel, just
as in the second, calibrated high resolution scan of
the first embodiment. In step 201, the high ~-esolution
image is spatially down-converted (decimated, filtered)
to a relatively low spatial resolution digitized iamge,
e.g. on the order of seven to twenty-four by ten to
thirty- six pixels per frame, so as to reduce the
computational intensity of the application of the high
spatial resolution image to the scene balance mechanism
through which that image is to be mapped into the
framestore.
This very low resolution (e.g. 24 x 36 pixel
sub-array) digitized image is then analyzed in step 202
by the scene balance mechanism as in the first
embodiment. Again, the scene balance mechanism outputs
three values, one for each color, which represent the
color balance content of the digitized image.
In step 203, using these values, the scene
balance mapping function is calibrated (shifted), so
that, during its application to the originally derived
high resolution image, the essential subject matter of
the image will be mapped in accordance with the linear
portion of the response range of the scanner's imaging
pixel array. Again, a respective offset code, one for
each of the color values, may be added to the inputs of
each scene balance look-up table in order to
effectively shift or translate its mapping function.
With the scene balance look~up tables shifted
on the basis of the analysis of the low resolution
converted image in step 203, the high resolution
digitized image is mapped into the framestore, in step
20~, such that essential image information (i.e. that
12~
which is necessary to obtain a high cluality print) is
captured and stored.
As will be appreciated from the foregoing
description, by subjecting a low resolution prescan of
an image to be scanned by a high resolution digital
scanner, the present invention is able to successfully
ensure that the essential subject matter of the image
will fall within the linear portion of the response
range of the scanner's imaging pixel array during a
high resolution scan. The high resolution digitized
image is then processed by the scene balance mechanism
to map the image data into a digitized image having a
reduced encoding resolution corresponding to that of an
attendant framestore. Although some of the information
in the captured scene is lost by the reduction in
encoding resolution, since the scanner has been 'scene
balance' calibrated, essential information (necessary ~ -
to reproduce a high cluality image) will be captured and
stored.
While I have shown and described an
embodiment in accordance with the present invention, it
is to be understood that the same is not limited
thereto but is susceptible to numerous changes and
modifications as known to a person skilled in the art,
and I therefore do not wish to be limited to the
details shown and described herein but intend to cover
all such changes and modifications as are obvious to
one of ordinary skill in the art.
