Facsimile and static presentation processing – Natural color facsimile – Scanning
Reexamination Certificate
1999-10-07
2003-09-09
Coles, Edward (Department: 2622)
Facsimile and static presentation processing
Natural color facsimile
Scanning
C358S533000, C358S445000, C358S446000, C358S447000, C358S483000, C250S208100, C382S274000, C382S275000, C382S254000
Reexamination Certificate
active
06618173
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of Invention
The present invention relates generally to a method for automatic prevention of vertical streaks and, more specifically, to a method for automatic prevention of vertical streaks by selectively applying gains to the output signals of optical sensor elements.
2. Description of the Related Art
Scanners typically include an array of optical sensor elements and a scan area (e.g., plate of glass) where an object to be imaged by the sensor elements is positioned. An optical path including, for example, lens and mirrors, spans between the sensor elements and the scan area.
Referring to
FIG. 2
, a subsystem
200
of a typical scanner includes an optical sensor device
202
, a lens
204
, a transparent plate
206
and a calibration strip
208
. The transparent plate
206
includes a scan area surface
210
over which an object
212
to be scanned is positioned. An exemplary calibration strip
208
spans across the entire scan area surface
210
and is formed from plastic with a uniform exterior color such as white.
The optical sensor device
202
is typically a linear array of optical sensor elements or photosites which convert optical images to electrical output signals. An exemplary optical sensor device
202
comprises a 2,700-bit×3CCD (Charge Coupled Device) color linear image sensor such as the NEC &mgr;PD3720 integrated circuit which has a color filter that provides primary colors (red, green and blue) via rows of photosites
214
,
216
and
218
, respectively, which are arranged on the sensor device
202
as shown.
A problem with the subsystem
200
is that different photosites, due to manufacturing imperfections, do not necessarily generate the same output signal when imaging identical objects. Another problem with the subsystem
200
If is that the optical path
220
(shown unfolded) between the optical sensor device
202
and the object
212
introduces inconsistencies in the output signals because the photosites at the end portions
222
and
224
of the optical sensor device
202
receive lower levels of light from an object
212
of uniform color than the photosites near the center portion
226
of the optical sensor device
202
. Therefore, in order to achieve uniformity in the levels of the output signals across the optical sensor device
202
, some form of compensation or calibration of the output signals is necessary. To this end, the subsystem
200
includes the calibration strip
208
which is used to calibrate the output signals of the optical sensor device
202
.
Referring to
FIG. 3
, a functional block diagram
300
shows that output signals
302
generated by the optical sensors
202
are provided with pixel-by-pixel gain
304
to generate calibrated output signals
304
. During the calibration process, the photosites of the optical sensor device
202
image the uniformly colored calibration strip
208
before the object
212
to be scanned is positioned on the scan area surface
210
. Each photosite in the scanner is “queried” to determine how much light it “sees”. Across the optical sensor device
202
, from the left end
222
to the right end
224
, the output signals
302
appear, for example, as shown in FIG.
4
. In order to achieve uniformity in the levels of the output signals across the optical sensor device
202
, a “proportionate” pixel-by-pixel gain
304
as shown in
FIG. 5
is applied to the output signals
302
. The term “proportionate” means an inversion or other appropriate function of the output signals
302
such that the calibrated output signals
304
appear as the uniform output level shown in FIG.
6
. By way of example, suppose an average photosite reports a value of 100. If one photosite reports a lower value—say 50—then the amplification for that one photosite will be set twice as high as the amplification for the average photosite. After the calibration process is completed, the pixel-by-pixel gain
304
is saved, for example, in firmware of the scanner, and applied during subsequent scanning. Thus, the net signal from the photosite and its amplification are the same for all photosite-amplification pairs.
Even though each photosite gets a “customized” amplification, unfortunately, this does not accommodate a situation where an optical obstruction is positioned between the calibration strip
208
and the scan area surface
210
during the calibration process. The term “optical obstruction” means an object which has any effect on light transmitted therethrough. Optical obstructions include, but are not limited to, paper dust, plastic dust, skin particles, metal particles and glass particles.
Referring again to
FIG. 2
, the subsystem
200
is shown with optical obstructions “A”, “B”, “C” and “D” positioned between the optical sensor device
202
and the calibration strip
208
. More specifically, the optical obstructions “A”, “B”, “C” and “D” are positioned, respectively, on the scan area surface
210
, in the optical path
220
, in the optical path
220
sufficiently near the scan area surface
210
to be illuminated by a light source (not shown), and on the optical sensor device
202
. The optical obstructions “A”, “B” and “D” are dark debris which are light-absorbing, i.e., tending to absorb light. The optical obstruction “C” is reflective. During the calibration process, when these optical obstructions are present, the output signals
302
, from the left end
222
to the right end
224
of the optical sensor device
202
, appear, for example, as shown in FIG.
7
. In order to achieve uniformity in the levels of the output signals across the optical sensor device
202
, a “proportionate” pixel-by-pixel gain
304
as shown in
FIG. 8
is applied to the output signals
302
. As shown in
FIG. 9
, a uniform photosite output signal level with proportionate gain applied is the result of the calibration process. However, if the optical obstruction “A” is displaced from the optical path
220
, for example, by an object
212
moving across the scan area surface
210
, the calibrated output signal levels will then appear as shown in
FIG. 10
with a large spike corresponding to the photosite that was imaging the optical obstruction “A” during the calibration process. As a result, during scanning, this erroneously high gain causes all scan data from that photosite to have a higher signal than it should. The net effect is that there is a bright vertical line in the scan, copy or fax output which runs the entire length of the image.
One possible approach to solving the problem described above would be to move the optical sensor device
202
relative to the calibration strip
208
during the calibration process. A disadvantage of such an approach is that it adds to the complexity of the scanner and makes it more expensive by requiring a mechanism for moving either the optical sensor device
202
or the calibration strip
208
relative to the other.
Thus, a need exists for a method for eliminating vertical streaks in scan data caused by optical obstructions in the optical path of the scanner. Also, a need exists for a method for “intelligently” determining when the signal from a photosite is truly low and needs to be compensated for with a large amplification and when a photosite is low due to dust which is likely to be dislodged from the optical path by the object being scanned and, therefore, needs to have an “ordinary” amplification.
SUMMARY OF THE INVENTION
A method for automatic prevention of vertical streaks in accordance with one embodiment of the present invention includes the steps of: processing output signals from an optical sensor; and selectively applying gains to the output signals depending upon differences between each output signal and its respective neighbor output signals.
In a preferred embodiment, the step of applying gains further includes identifying output signals to which a proportionate gain is to be applied and output signals to which a gain which is appropriate for at least one of the neighbor output signals is to be applied. Output signals to which proportionate
Nobel Gary M.
Wee Daniel
Coles Edward
Hewlett-Packard Development Company LP.
Safaipour Houshang
LandOfFree
Method for automatic prevention of vertical streaks by... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for automatic prevention of vertical streaks by..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for automatic prevention of vertical streaks by... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3109036