Title: Digitizer with improved dynamic range and photometric resolution
Abstract: A digitizer having a dual exposure technique is combined with an associated LUT for each exposure. Each LUT may have a transfer function including a logarithmic operator resulting in a digitized image with improved photometric resolution and increased dynamic range. A digitizer utilizing multiple exposures at approximately equal exposure time intervals provides a noise reduction for lower optical density portions of the data medium further contributing to increased dynamic range. Associated methods are also provided.
Patent Number: 6,914,701 Issued on 07/05/2005 to Lehman
| Inventors:
|
Lehman; Richard (Nashua, NH)
|
| Assignee:
|
Howtek Devices Corporation (Hudson, NH)
|
| Appl. No.:
|
313764 |
| Filed:
|
December 6, 2002 |
| Current U.S. Class: |
358/445; 358/487; 358/443; 358/448; 358/463; 358/482; 358/483; 358/506; 358/444; 358/475; 358/513; 358/514; 358/523; 341/155; 341/156; 341/172 |
| Intern'l Class: |
H04N 001/40; H04N001/04; H04N001/38; H03M001/12 |
| Field of Search: |
358/487,506,445,443,463,448,444,475,513,523,446,509,497,494,496,498,483,482
341/155,156,172,139-141
382/274,275,312,318,319
250/208.1,234-236
|
References Cited [Referenced By]
U.S. Patent Documents
| 3952290 | Apr., 1976 | Williams.
| |
| 4314281 | Feb., 1982 | Wiggins et al.
| |
| 4396903 | Aug., 1983 | Habicht et al.
| |
| 4408231 | Oct., 1983 | Bushaw et al.
| |
| 4439788 | Mar., 1984 | Frame.
| |
| 4485405 | Nov., 1984 | Bailey.
| |
| 4540901 | Sep., 1985 | Suzuki.
| |
| 4647975 | Mar., 1987 | Alston et al.
| |
| 4654713 | Mar., 1987 | Boucharlat et al.
| |
| 4665440 | May., 1987 | Tromborg.
| |
| 4670787 | Jun., 1987 | Levine.
| |
| 4691237 | Sep., 1987 | Shimizu.
| |
| 4772958 | Sep., 1988 | Suzuki.
| |
| 4878119 | Oct., 1989 | Beikirch et al.
| |
| 4884224 | Nov., 1989 | Hirosawa.
| |
| 4992653 | Feb., 1991 | Kawahara et al.
| |
| 5034825 | Jul., 1991 | Ejiri et al.
| |
| 5093734 | Mar., 1992 | Gerlach.
| |
| 5099359 | Mar., 1992 | Hrycin et al.
| |
| 5101108 | Mar., 1992 | Gaalema et al.
| |
| 5192856 | Mar., 1993 | Schaham.
| |
| 5218485 | Jun., 1993 | Malm.
| |
| 5221848 | Jun., 1993 | Milch.
| |
| 5241404 | Aug., 1993 | Furukawa et al.
| |
| 5251406 | Oct., 1993 | Kirn.
| |
| 5252818 | Oct., 1993 | Gerlach et al.
| |
| 5268567 | Dec., 1993 | Bowlby, Jr. et al.
| |
| 5343308 | Aug., 1994 | Johnson.
| |
| 5410144 | Apr., 1995 | Lavelle et al.
| |
| 5416611 | May., 1995 | Tandon.
| |
| 5422643 | Jun., 1995 | Chu et al.
| |
| 5682033 | Oct., 1997 | Cattorini.
| |
| 5955725 | Sep., 1999 | Cattorini.
| |
| 6271940 | Aug., 2001 | Deschuytere et al.
| |
| 6404516 | Jun., 2002 | Edgar.
| |
Primary Examiner: Lee; Cheukfan
Attorney, Agent or Firm: Grossman Tucker Perreault & Pfleger, PLLC
Claims
1. A digitizer for digitizing an image on a data medium, said digitizer comprising:
a light sensitive element configured to be responsive to a light beam from an
illuminator during a first exposure and during a second exposure, said first exposure
and said second exposure being produced by controlling a time that said light sensitive
element is responsive to said light beam, said first exposure associated with a
first range of optical densities and said second exposure associated with a second
range of optical densities of said image, said light sensitive element configured
to produce a first set of analog signals associated with said first exposure and
a second set of analog signals associated with said second exposure;
an analog to digital converter configured to convert said first set of analog
signals to a first set of digital signals and said second set of analog signals
to a second set of digital signals, said first set of analog signals being produced
by said light sensitive element during a first time interval that said light sensitive
element is sensitive to said light beam, and said second set of analog signals
being produced by said light sensitive element during a second time interval that
said light sensitive element is sensitive to said light beam, wherein said second
time interval is greater than said first time interval by a multiplication factor
n, and wherein a boundary optical density between said first range of optical densities
and said second range of optical densities is equal to Log(n); and
a machine-readable medium comprising a first look up table and a second look
up table, wherein said first look up table is configured to map said first set
of digital signals to a first set of output signals associated with said first
range of optical densities, and said second look up table is configured to map
said second set of digital signals to a second set of output signals associated
with said second range of optical densities.
2. The digitizer of claim 1, wherein said multiplication factor n is 20 and wherein
said boundary optical density is 1.3.
3. A digitizer for digitizing an image on a data medium, said digitizer comprising:
a light sensitive element configured to be responsive to a light beam from an
illuminator during a first exposure and during a second exposure, said first exposure
associated with a first range of optical densities and said second exposure associated
with a second range of optical densities of said image, said light sensitive element
configured to produce a first set of analog signals associated with said first
exposure and a second set of analog signals associated with said second exposure
an analog to digital converter configured to convert said first set of analog
signals to a first set of digital signals and said second set of analog signals
to a second set of digital signals; and
a machine-readable medium comprising a first look up table and a second look
up table, wherein said first look up table is configured to map said first set
of digital signals to a first set of output signals associated with said first
range of optical densities, and said second look un table is configured to map
said second set of digital signals to a second set of output signals associated
with said second range of optical densities, wherein said first look up table has
a first transfer function for mapping said first set of digital signals to said
first set of output signals, and wherein said second look up table has a second
transfer function for mapping said second set of digital signals to said second
set of output signals, wherein said first transfer function and said second transfer
function comprise a logarithmic operator.
4. The digitizer of claim 3, wherein said first set of digital signals and said
second set of digital signals are 16-bit signals and wherein said first set of
output signals and said second set of output signals are 12-bit signals.
5. The digitizer of claim 4, wherein said first transfer function and second
transfer function are given by the equation:
wherein said output comprises one of said first set and said second set of output
signals; said Maxlog comprises a predetermined maximum count value for said first
set or said second set of output signals; said Maxim comprises a predetermined
maximum count value for said first set or said second set of digital signals; and
said Count comprises one of said first set and said second set of digital signals.
6. The digitizer of claim 5, wherein said Maxlog is 4,000 and said Maxim is 64,000.
7. A digitizer with noise reduction for low density images, said digitizer comprising:
an illuminator configured to generate a light beam to illuminate a data medium;
a light sensitive element configured to be sensitive to said light beam for a plurality
of approximately equal time intervals, said light sensitive element further configured
to accept said light beam and convert said light beam into sets of analog signals
associated with each of said plurality of time intervals representing data recorded
on the data medium; an analog to digital converter configured to convert each of
said sets of analog signals to an associated set of digital signals, wherein each
of said sets of digital signals has an associated noise level; and
a processor configured to average each of said sets of digital signals into an
average digital signal having count values, said average digital signal having
a second noise level, said second noise level being less than said each said associated
noise level for each said set of digital signals.
8. The digitizer of claim 7, wherein said second noise level is less than said
associated noise level for each said set of digital signals based on a factor dependent
on a number of times said data medium is exposed to said light from said light
source for said approximately equal time intervals.
9. A method of presenting data from a dual exposure technique in a digitizer,
said method comprising the steps of:
exposing a data medium to a first exposure associated with a first range of optical
densities and to a second exposure associated with a second range of optical densities;
accumulating a first set of analog charges associated with said first exposure
and a second set of analog charges associated with said second exposure;
converting said first set of analogs charges to a first set of digital signals
and converting said second set of analog charges to a second set of signals; and
mapping said first set of digital signals to a first set of output signals and
said second set of digital signals to a second set of output signals, wherein said
mapping step comprises the step of applying a logarithmic operator.
10. A method of presenting data from a dual exposure technique in a digitizer,
said method comprising the steps of:
exposing a data medium to a first exposure associated with a first range of optical
densities and to a second exposure associated with a second range of optical densities;
accumulating a first set of analog charges associated with said first exposure
and a second set of analog charges associated with said second exposure;
converting said first set of analog charges to a first set of digital signals
and converting said second set of analog charges to a second set of signals; and
mapping said first set of digital signals to a first set of output signals and
said second set of digital signals to a second set of output signals, wherein said
first set of digital signals and said second set of digital signals are 16-bit
signals and wherein said first set of output signals and said second set of output
signals are 12-bit signals.
11. The method of claim 10, wherein said mapping step comprises application of
a transfer function given by the equation:
wherein said output comprises one of said first set and said second set of output
signals; said Maxlog comprises a predetermined maximum count value for said first
set or said second set of output signals; said Maxim comprises a predetermined
maximum count value for said first set or said second set of digital signals; and
said Count comprises one of said first set and said second set of digital signals.
12. The method of claim 11, wherein said Maxlog is 4,000 and said Maxim is 64,000.
13. A method of improving photometric resolution from a dual exposure technique
in a digitizer, said method comprising the steps of:
exposing a data medium to a light beam for a first exposure time interval;
accumulating a first set of analog charges associated with said first exposure
time interval;
exposing said data medium to said light beam for a second exposure time interval,
wherein said second exposure time interval is greater than said first exposure
time interval by a multiplication factor n;
accumulating a second set of analog charges associated with said second exposure
time interval;
converting said first set of analog charges to a first set of linear digital
count values and converting said second set of analog charges to a second set of
linear digital count values;
mapping said first set of linear digital count values corresponding to portions
of said data medium having an optical density less than or equal to Log n to a
first set of logarithmic digital count values; and
mapping said second set of linear digital count values corresponding to portions
of said data medium having an optical density greater than Log n to a second set
of logarithmic digital count values.
14. A method of reducing noise for low optical density portions of a data medium,
said method comprising the steps of:
exposing said data medium to a light beam for a plurality of substantially equal
exposure time intervals;
accumulating a set of analog charges associated with each of said plurality of
substantially equal exposure time intervals;
convertering each said set of analog charges to an associated set of digital
count values each having an associated noise level; and
averaging each said set of digital count values to an average digital representation
having a second associated noise level, wherein said second associated noise level
is less than said associated noise level for each said set of digital count values.
Description
FIELD OF THE INVENTION
The present invention relates generally to a digitizer and in particular to a
digitizer with improved dynamic range and photometric resolution.
BACKGROUND OF THE INVENTION
In general, digitizers convert images on various media to an electric signal
which
can then be stored, transferred, or analyzed in any number of ways. The image captured
on the media can be described by a two-dimensional array of picture elements or
pixels quantified in terms of the transmittance or optical density of the medium
at the particular coordinates of the pixels.
A medium that has regions of high optical density (low transmittance) and low
optical
density (high transmittance) requires a digitizer capable of accurately reading
such image data. Some media, e.g., transparent media such as X-ray films, have
images with such a wide range of optical densities. One way of measuring the performance
of a digitizer system to capture such a wide range of image data is its dynamic
range. Dynamic range is generally defined as the ratio of the maximum output signal
of a light detector of the digitizer when illuminated with light and the noise
output in the absence of light. It is typically expressed as the Log (White signal/RMS
noise). Any reduction in noise would therefore serve to effectively increase the
dynamic range of the digitizer.
Accordingly, there is a need in the art for a digitizer capable of reducing
noise and therefore improving dynamic range, as well as increasing the photometric
resolution of a digitized image.
BRIEF SUMMARY OF THE INVENTION
A digitizer for digitizing an image on a data medium consistent with the invention
includes: a light sensitive element configured to be responsive to a light beam
from an illuminator during a first exposure and during a second exposure, the first
exposure associated with a first range of optical densities and the second exposure
associated with a second range of optical densities of the image, the light sensitive
element configured to produce a first set of analog signals associated with the
first exposure and a second set of analog signals associated with the second exposure;
an analog to digital converter configured to convert the first set of analog signals
to a first set of digital signals and the second set of analog signals to a second
set of digital signals; and a machine-readable medium includes a first look up
table and a second look up table. The first look up table is configured to map
the first set of digital signals to a first set of output signals associated with
the first range of optical densities, and the second look up table is configured
to map the second set of digital signals to a second set of output signals associated
with the second range of optical densities.
According to another aspect of the invention, there is provided a digitizer
with noise reduction for low density images including: an illuminator configured
to generate a light beam to illuminate a data medium; a light sensitive element
configured to be sensitive to the light beam for a plurality of approximately equal
time intervals, the light sensitive element further configured to accept the light
beam and convert the light beam into sets of analog signals associated with each
of the plurality of time intervals representing data recorded on a data medium;
an analog to digital converter configured to convert each set of analog signals
to an associated set of digital signals, wherein each of set digital signals has
an associated noise level; and a processor configured to average each set of digital
signals into an average digital signal having count values, the average digital
signal having a second noise level, the second noise level less than each associated
noise level for each set of digital signals.
According to a further aspect of the invention, there is provided a method
of presenting data from a dual exposure technique in a digitizer including the
steps of: exposing a data medium to a first exposure associated with a first range
of optical densities and to a second exposure associated with a second range of
optical densities; accumulating a first set of analog charges associated with the
first exposure and a second set of analog charges associated with the second exposure;
converting the first set of analog charges to a first set of digital signals and
converting the second set of analog charges to a second set of signals; and mapping
the first set of digital signals to a first set of output signals and the second
set of digital signals to a second set of output signals.
According to a further aspect of the invention, there is provided a method
of improving photometric resolution from a dual exposure technique in a digitizer
including the steps of: exposing a data medium to a light beam for a first exposure
time interval; accumulating a first set of analog charges associated with the first
exposure time interval; exposing the data medium to the light beam for a second
exposure time interval, wherein the second exposure time interval is greater than
the first exposure time interval by a multiplication factor n; accumulating a second
set of analog charges associated with the second exposure time interval; converting
the first set of analog charges to a first set of linear digital count values and
converting the second set of analog charges to a second set of linear digital count
values; mapping the first set of linear digital count values corresponding to portions
of the data medium having an optical density less than or equal to Log n to a first
set of logarithmic digital count values; and mapping the second set of linear digital
count values corresponding to portions of the data medium having an optical density
greater than Log n to a second set of logarithmic digital count values.
According to yet a further aspect of the invention, there is provided a
method of reducing noise for low optical density portions of a data medium including
the steps of: exposing the data medium to a light beam for a plurality of substantially
equal exposure time intervals; accumulating a set of analog charges associated
with each plurality of substantially equal exposure time intervals; convertering
each set of analog charges to an associated set of digital count values each having
an associated noise level; and averaging each set of digital count values to an
average digital representation having a second associated noise level, wherein
the second associated noise level is less than the associated noise level for each
set of digital count values.
BRIEF DESCRIPTION OF THE DRAWINGS
Advantages of the present invention will be apparent from the following
detailed description of exemplary embodiments thereof, which description should
be considered in conjunction with the accompanying drawings, in which:
FIG. 1 is a simplified block diagram of a digitizer system consistent with the
invention having a digitizer portion and host computer portion;
FIG. 2A is an exemplary table of linear count values associated with a range
of optical densities for a first exposure time interval of 1×;
FIG. 2B is an exemplary table of linear count values associated with a range
of optical densities for a second exposure time interval of 20×;
FIG. 3A is an exemplary first lookup table for mapping the linear count data
from FIG. 2A to logarithmic count data; and
FIG. 3B is an exemplary second lookup table for mapping the linear count data
from FIG. 2B to logarithmic count data.
DETAILED DESCRIPTION
FIG. 1 illustrates a simplified block diagram of a digitizer system
100a
consistent with the invention having a digitizer portion
100 and a host
computer portion
140. The digitizer
100 generally includes an illuminator
101, a lens
104, a light sensitive element
106, and electrical
processing circuitry
114 including an analog to digital converter
108
and a processor
112. The digitizer
100 may also include a machine-readable
medium
131 for storing digital data such as a first lookup table
132
and a second lookup table
134 whose operation is later detailed. In general,
the digitizer
100 converts optical densities of an image on a data medium
102 to electrical signals that represent various intensities of transmitted
light and thus densities of the data on the applicable medium.
The data medium
102 is first located in a position
102′
to be digitized. The digitizer
100 may include a transport control system
(not shown) to accept the data medium
102 and drive it to the position
102′
to be scanned. Alternatively, the digitizer
100 may be a flat bed type scanner
where a user would position the data medium
102 in position
102′
and the illuminator
101 and light sensitive element
106 would move
relative to the data medium
102 while in position
102′.
The data medium
102 may be any type of medium, e.g., transparent media
or reflective media. The digitizer
100 is illustrated as a digitizer for
digitizing transparent media since the illuminator
101 is positioned above
the object plane. However, a digitizer for digitizing reflective media may also
be utilized where the illuminator would be positioned below the object plane.
Once the data medium
102 is properly positioned, a portion of light from
the illuminator
101 passes through the data medium
102. The amount
of light passing through the data medium depends on the transmittance or optical
density of the particular image on the data medium
102 at each pixel. A
lens
104 may also be used to image the light onto the light sensitive element
106.
The light sensitive element
106 accumulates photons and converts such
photons into an analog electrical signal representative of the accumulated photons.
The light sensitive element
106 may by a variety of elements known in the
art such as a charge coupled device (CCD) array or a CMOS array. The light sensitive
element
106 may be in the form of a line, square, rectangle, or any various
shape such that the whole data medium
102 may be virtually divided into
areas and every area corresponds to an analog signal that represents the amount
of light transmitted through that area. In this manner, an analog image signal
may be obtained which represents the whole image recorded on the data medium
102.
The analog image is then processed by electronic processing circuitry
114.
The electronic processing circuitry
114 may include a variety of devices
known in the art including the analog to digital converter
108 for converting
the analog signal into a digital signal and a processor
112. As such, a
digital image signal may be output to terminal
116 for further electronic
use, e.g., electronic storage, processing, and communication. A host computer
140
having a variety of components known to those skilled in the art may also be coupled
to the output terminal
116. Such components may include a display monitor
142 for displaying digitized data and machine-readable storage
144
for storing digital data, and its own CPU
146.
When digitizing a data medium
102 having image data with a wide range
of optical densities, each line scan of the data medium
102 may be exposed
to a plurality of exposures and a look up table (LUT) associated with each exposure
may then be utilized as further detailed herein. Description is made to two separate
exposures and two separate LUTs
132,
134, although any plurality
of exposure and associated LUTs may be utilized in a digitizer consistent with
the present invention. In addition, description is made herein to achieve each
exposure by varying the amount of time that the light sensitive element accumulates
photons. Exposures could also be made by varying the light intensity of the illuminator
101 or by other methods known in the art.
When each line of the data medium
102 is being digitized, the light sensitive
element
106 may accumulate photons until an appropriate control signal instructs
the light sensitive element
106 to stop accumulating photons after a predetermined
time interval. Such a control signal may be provided by a variety of components
known in the art such as the processor
112.
Accumulation of photons in the light sensitive element
106 may
therefore occur during a first exposure time interval and a second exposure time
interval, where the second time interval is greater than the first time interval,
e.g., by a multiplication factor n. Since the second time interval is greater than
the first time interval, such exposure is directed at those images recorded on
the data medium that have a higher range of optical densities. In contrast, the
first exposure is directed at those images recorded on the data medium that have
a lower range of optical densities. The boundary optical density between the first
lower optical density range and the higher optical density range is the Log of
the multiplication factor n.
For instance, in one embodiment that multiplication factor n is 20 corresponding
to a 1× exposure for the first exposure and a 20× exposure for the second
exposure. The Log
20 is 1.3 such that an optical density of 1.3 is the boundary
between the first lower optical density range and the second higher optical density
range. With an approximate density range of 0.0 to 4.0 for the digitizer
100,
the first exposure time is directed at optical densities between 0.0 and 1.3 and
the second exposure time is directed at optical densities between 1.3 and 4.0.
The actual time for each exposure depends on a number of factors including the
characteristics of the light sensitive element
106. For instance, the 1×
integration time could be on the order of 200 microseconds and the 20× integration
time could be on the order of 4,000 microseconds.
Turning to FIG. 2A, an exemplary table
200A of data illustrating the
various voltages and linear counts that would be obtained at various transmittance
and optical densities for the first exposure (1×) is illustrated. Transmittance
levels and corresponding density levels for each transmittance level are illustrated
in the first two columns. The associated analog voltage and corresponding digital
count value are illustrated in the next two columns.
The maximum analog voltage signal is 2.0 volts in this example, which is present
if the transmittance is 1.0. The analog voltage signal is linearly reduced as the
transmittance is reduced. For instance, at a transmittance of 0.05 or an OD of
1.3, the voltage signal is 0.1 volts. Assuming the analog to digital converter
is a 16-bit converter, the maximum count value would be 65,536. In the exemplary
table of FIG. 2A, the maximum count value was established at 64,000 for convenience.
The count value would also decrease linearly with the reduction in the analog voltage signal.
Turning to FIG. 2B, an exemplary table
200B of data illustrating the
various voltages and linear counts that would be obtained at various transmittance
and optical densities for the second exposure (20×) is illustrated. Given
the longer exposure time, the analog voltage level is saturated at 2.0 volts for
those optical densities less than the boundary optical density level of 1.3 in
this example. For optical densities greater than 1.3, the analog voltage signal
is not saturated and the linear count data, assuming a 16-bit A/D converter and
establishing the maximum count value of 64,000, ranges from 64,000 down to near zero.
Since there are now two sets of linear count data associated with each exposure,
the electronic processing circuitry
114 must take the appropriate linear
count data for each respective optical density level for each pixel. As illustrated
in the exemplary tables
200A and
200B, the electronic processing
circuitry does not have to make any comparison or selection amongst the linear
count data. Rather, the electronic processing circuitry simply takes the non-saturated
count data from the second exposure (corresponding to optical densities greater
than 1.3 in this example) and takes the 1× linear count data for all other
optical densities. As such, the darker image data at higher optical densities has
linear count data based upon the longer exposure time interval.
Given this dual exposure technique, there is a greater amount of linear count
data representing the higher optical density range between densities of 1.3 and
4.0. For instance, there are 64,000 counts of data for the second exposure as opposed
to only 3,200 counts for the first exposure for the optical density range between
1.3 and 4.0. Corresponding LUTs for each exposure may then be utilized together
with the appropriate linear count data from each exposure in order to map input
count data to output count data. Each LUT may map linear input data to linear output
data to produce a low noise linear output signal. Alternatively, each LUT may have
a transfer function that includes a logarithmic operator in order to improve photometric
resolution of the digitizer. Such exemplary LUTs
132,
134 including
a logarithmic operator are further detailed herein.
The LUTs
132,
134 may be stored in any variety of machine-readable
media
131, e.g., random access memory (RAM), read only memory (ROM), magnetic
disk (e.g., floppy disk or hard disk drive), optical disk (e.g., CD/DVD ROM), and
any other device that can store digital information. The machine-readable media
131 is part of the digitizer system
100a which may be included
in the digitizer
100 as illustrated in FIG. 1 or in the host computer
140.
Details of an exemplary first LUT
130 are illustrated in FIG.
3A.
In this exemplary LUT, the input data is 16-bit linear data with an established
maximum linear count value of 64,000 as illustrated in the first column of the
LUT. The output column of the LUT in this example is 12-bit log count data normalized
to a maximum log count value of 4,000. Those skilled in the art will recognize
that a variety of linear input levels and output levels may be utilized depending
on the size of the A/D converter and the chosen maximum count values.
In the exemplary LUT
130 of FIG. 3A, the 12-bit log output data is given
by the transfer function:
The Maxlog value represents the selected maximum log count value depending on
the number of bits needed. With a 12-bit output, the maximum count value would
be 4,096 and 4,000 was selected as Maxlog to obtain the output values illustrated
in FIG.
3A. Similarly, the Maxlin value is selected based on the size of
A/D converter and the selected maximum value in this instance was 64,000 as previously
detailed. The Count value represents the linear count value of the input. The exemplary
LUT of FIG. 3
a thus maps input linear count data for those optical densities
between 0.0 and 1.3 to associated log count output data. Since Log (Maxlin/Count)
is equal to density, and Maxlog is 4,000 in this example, equation (1) may be simplified
to equation (1a) below:
In addition, the transfer function of equation (1) may be amended to include a
bias offset that is subtracted from the Maxlin and Count values. The bias offset
is an intentional offset to avoid having the analog signal be a negative number
since such a negative number may not be properly converted by an A/D converter.
Turning to FIG. 3B, another exemplary LUT
134 is illustrated for mapping
linear count data for those optical densities between 1.3 and 4.0 in this example
to output values. The LUT
134 is similar to the earlier described LUT
132
and the output data of the LUT
134 may be defined by a transfer function
as detailed in equation 1. Similar to the first LUT
132, the transfer function
for the second LUT
134 may be amended to include a bias offset that is subtracted
from the Maxlin and Count values. A factor to account for actual exposure times
may also be included and would be multiplied by the Maxlin value.
The dual exposure technique in combination with the appropriate dual LUTs enables
photometric resolution of the digitized image to be improved. For instance, there
are only 3,200 linear counts of data over the 2.7 density range between 1.3 and
4.0 resulting from the 1× exposure. In contrast, there are 64,000 counts of
data from the second exposure for the same range of optical densities. As such,
the second LUT
134 can utilize these 64,000 counts of data to improve photometric
resolution over this higher optical density range since the brightness of the resulting
digitized image is dependent on the number of count values.
The dual exposure and dual LUT technique also effectively serves to increase
dynamic range of the digitizer by extending capabilities of the digitizer in the
higher optical density range. The improvement in dynamic range is dependent on
the multiplication factor n or the length of overexposure for the second exposure
period. For example, when n is equal to 20, densities higher than 1.3 will have
noise reduced by 95% resulting in a 1.3 density improvement in dynamic range if
noise is a function of black noise only. When n is equal to 10, densities higher
than 1.0 will have noise reduced by 90% resulting in a 1.0 density improvement
in dynamic range if noise is a function of black noise only.
In order to reduce noise in lower optical density ranges, the multiplication
factor
n can be lowered thus effectively reducing the length of overexposure. For instance,
a multiplication factor of 2 would result in a noise improvement for optical densities
above 0.3 as opposed to noise improvements for optical densities above 1.3 when
the multiplication factor is 20.
In addition, multiple sampling at 1× exposure can further reduce noise at
lower optical density values and thus further improve dynamic range of the digitizer.
The reduction in noise is dependent on the number of times the image can be sampled
at 1× exposure. For instance, if the data medium
102 was exposed to
three different exposures at a 1× time interval and the results for each exposure
were averaged, the noise could be reduced for all optical densities by a factor
based on the square root of 3 or an approximate noise reduction of 43%. Four samplings
at 1× would result in a 50% reduction in noise. Ten samplings at 1× would
result in a 68% reduction in noise.
The amount of 1× sampling should be balanced with the increased time it
will take for digitizing the data medium. As light sensitive elements become more
sensitive, they can be exposed to light for shorter amounts of time and hence noise
can be lowered without unduly increasing scan times.
The embodiments that have been described herein, however, are but some of the
several which utilize this invention and are set forth here by way of illustration
but not of limitation. It is obvious that many other embodiments, which will be
readily apparent to those skilled in the art, may be made without departing materially
from the spirit and scope of the invention as defined in the appended claims.
*
