Patent Application
20080292182
The invention relates generally to the field of digital image processing operations that produce a full-color noise-reduced full-resolution image from an image having panchromatic and color pixels.
One of the most common and frequently essential image processing operations is noise reduction. This is especially true for digital still and video camera images that may have been captured under insufficient lighting conditions. One way to address digital image capture under less than optimum lighting conditions is to either acquire or synthesize one or more color channels that are particularly sensitive to low or insufficient scene illumination. The data from the channels with increased light sensitivity are generally used to guide the subsequent image processing of the data from the accompanying standard color channels. Noise reduction is a prime candidate for benefiting from this additional image data. A number of examples exist in the literature. U.S. Pat. No. 6,646,246 (Gindele, et al.) teaches using an extended dynamic range color filter array (CFA) pattern with slow and fast pixels, noise-cleaning the slow pixel data using only slow pixel data and noise-cleaning the fast pixel data using only fast pixel data. This approach achieved noise reduction at the expense of image resolution as each color channel is now subdivided into a fast channel and a slow channel and the subsequent merger can produce image processing artifacts more troublesome than the addressed original noise. U.S. Pat. No. 7,065,246 (Xiaomang, et al.) is representative of a fair number of similarly disclosed inventions in that it reveals constructing a luminance signal from directly sensed color channel data, in this case cyan, magenta, yellow, and green. The high-frequency component of the constructed luminance is used to replace the high-frequency component of the original color channel signals to affect a net noise reduction of the image data. Although somewhat effective, the major liability of this approach is that the synthesized luminance channel is constructed from noisy color channel components resulting in an essentially equally noisy synthetic channel.
A suggestion of a better approach can be found in U.S. Pat. No. 5,264,924 (Cok). Cok discloses direct measurement of red, green, blue, and luminance values at each pixel location. The high-frequency luminance data which is designed to be inherently less noisy than the corresponding high-frequency red, green, and blue data is used to replace said high-frequency red, green, and blue data to produce noise-cleaned red, green, and blue signals. Since the vast majority of digital still and video cameras use a single sensor equipped with a CFA that only senses one color channel per pixel, Cok cannot be directly practiced in such systems.
Although Xiaomang and Cok describe luminance signals, a color channel with photometric sensitivity conforming to the luminance channel of the human visual system may be unnecessarily restrictive.
It is an object of the present invention to produce a noise-reduced full-resolution full-color image from a digital image having panchromatic and color pixels.
This object is achieved by a method for producing a noise-reduced digital color image, comprising:
a. providing an image having panchromatic pixels and color pixels corresponding to at least two color photoresponses;
b. providing from the image a panchromatic image and at least one color image; and
c. using the panchromatic image and the color image to produce the noise-reduced digital color image by setting a plurality of color characteristics equal to the corresponding panchromatic characteristics at each color pixel location.
It is a feature of the present invention that images can be captured under low-light conditions with a sensor having panchromatic and color pixels and processing reduces noise in a full-color image produced from the panchromatic and colored pixels. A more useful signal can be captured with a more general panchromatic channel, which has higher sensitivity at all wavelengths over the luminance channel of the human visual system.
In the following description, a preferred embodiment of the present invention will be described in terms that would ordinarily be implemented as a software program. Those skilled in the art will readily recognize that the equivalent of such software can also be constructed in hardware. Because image manipulation algorithms and systems are well known, the present description will be directed in particular to algorithms and systems forming part of, or cooperating more directly with, the system and method in accordance with the present invention. Other aspects of such algorithms and systems, and hardware or software for producing and otherwise processing the image signals involved therewith, not specifically shown or described herein, can be selected from such systems, algorithms, components and elements known in the art. Given the system as described according to the invention in the following materials, software not specifically shown, suggested or described herein that is useful for implementation of the invention is conventional and within the ordinary skill in such arts.
Still further, as used herein, the computer program can be stored in a computer readable storage medium, which can comprise, for example; magnetic storage media such as a magnetic disk (such as a hard drive or a floppy disk) or magnetic tape; optical storage media such as an optical disc, optical tape, or machine readable bar code; solid state electronic storage devices such as random access memory (RAM), or read only memory (ROM); or any other physical device or medium employed to store a computer program.
Before describing the present invention, it facilitates understanding to note that the present invention is preferably utilized on any well-known computer system, such as a personal computer. Consequently, the computer system will not be discussed in detail herein. It is also instructive to note that the images are either directly input into the computer system (for example by a digital camera) or digitized before input into the computer system (for example by scanning an original, such as a silver halide film).
Referring to
A compact disk-read only memory (CD-ROM) 124, which typically includes software programs, is inserted into the microprocessor based unit for providing a way of inputting the software programs and other information to the microprocessor based unit 112. In addition, a floppy disk 126 can also include a software program, and is inserted into the microprocessor-based unit 112 for inputting the software program. The compact disk-read only memory (CD-ROM) 124 or a floppy disk 126 can alternatively be inserted into an externally located disk drive unit 122 which is connected to the microprocessor-based unit 112. Still further, the microprocessor-based unit 112 can be programmed, as is well known in the art, for storing the software program internally. The microprocessor-based unit 112 can also have a network connection 127, such as a telephone line, to an external network, such as a local area network or the Internet. A printer 128 can also be connected to the microprocessor-based unit 112 for printing a hardcopy of the output from the computer system 110.
Images can also be displayed on the display 114 via a personal computer card (PC card) 130, such as, as it was formerly known, a PCMCIA card (based on the specifications of the Personal Computer Memory Card International Association) that contains digitized images electronically embodied in the PC card 130. The PC card 130 is ultimately inserted into the microprocessor-based unit 112 for permitting visual display of the image on the display 114. Alternatively, the PC card 130 can be inserted into an externally located PC card reader 132 connected to the microprocessor-based unit 112. Images can also be input via the compact disk 124, the floppy disk 126, or the network connection 127. Any images stored in the PC card 130, the floppy disk 126 or the compact disk 124, or input through the network connection 127, can have been obtained from a variety of sources, such as a digital camera (not shown) or a scanner (not shown). Images can also be input directly from a digital camera 134 via a camera docking port 136 connected to the microprocessor-based unit 112 or directly from the digital camera 134 via a cable connection 138 to the microprocessor-based unit 112 or via a wireless connection 140 to the microprocessor-based unit 112.
In accordance with the invention, the algorithm can be stored in any of the storage devices heretofore mentioned and applied to images in order to interpolate sparsely populated images.
Returning to
X
5 l =(
P
1
+P
2
+P
3
+P
7
+P
8
+P
9)/6
Alternate weighting to the pixel value in this approach are also well known to those skilled in the art. As an example,
X
5=(P1+2P2+P3+P7+2P8+P9)/8
Alternately, an adaptive approach can be used by first computing the absolute values of directional gradients (absolute directional gradients).
B
5
=|P
1
−P
9|
V
5
=|P
2
−P
8|
S
5
=|P
3
−P
7|
The value of X5 is now determined by one of three two-point averages.
BX
5=(P1+P9)/2
VX
5=(P2+P8)/2
SX
5=(P3+P7)/2
The two-point average associated with the smallest value of the set of absolute direction gradients is used for computing X5, e.g., if V5≦B5 and V5≦S5, then X5=VX5.
Returning to
R1−R5 (north-west)
R2−R5 (north)
R3−R5 (north-east)
R4−R5 (west)
R6−R5 (east)
R7−R5 (south-west)
R8−R5 (south)
R9−R5 (south-east).
Another example of a color characteristic is the second spatial pixel difference. Again referring to
2R5−R1−R9 (backslash)
2R5−R2−R8 (vertical)
2R5−R3−R7 (slash)
2R5−R4−R6 (horizontal).
Four examples of the invention are now given. Referring to
The weighting coefficients c1 through c9 define a valid pixel neighborhood and are computed from differences in panchromatic values and from differences in red values.
c
1=1 if |P1−P5|≦tp and |R1−R5|≦tr, otherwise c1=0
c
2=1 if |P2−P5|≦tp and |R2−R5|≦tr, otherwise c2=0
c
3=1 if |P3−P5|≦tp and |R3−R5|≦tr, otherwise c3=0
c
4=1 if |P4−P5|≦tp and |R4−R5|≦tr, otherwise c4=0
c
5=1 if |P5−P5|≦tp and |R5−R5|≦tr, otherwise c5=0
c
6=1 if |P6−P5|≦tp and |R6−R5|≦tr, otherwise c6=0
c
7=1 if |P7−P5|≦tp and |R7−R5|≦tr, otherwise c7=0
c
8=1 if |P8−P5−≦tp and |R8−R5|≦tr, otherwise c8=0
c
9=1 if |P9−P5|≦tp and |R9−R5|≦tr, otherwise c9=0
In these expressions tp and tr are predetermined positive threshold values that are chosen to exclude pixel values that are separated from the central pixel (R5) by any edges in the pixel neighborhood shown in
An alternate method to the one described above is to use the second spatial pixel difference averaged over a valid pixel neighborhood. Again referring to
The weighting coefficients d1 through d4 define a valid pixel neighborhood and are computed from differences in panchromatic values and from differences in red values.
d
1=1 if |P1−P5|≦tp and |P9−P5|≦tp and |R1−R5|≦tr and |R9−R5|≦tr, otherwise d1=0
d
2=1 if |P2−P5|≦tp and |P8−P5|≦tp and |R2−R5|≦tr and |R8−R5|≦tr, otherwise d2=0
d
3=1 if |P3−P5|≦tp and |P7−P5|≦tp and |R3−R5|≦tr and |R7−R5|≦tr, otherwise d3=0
d
4=1 if |P4−P5|≦tp and |P6−P5|≦tp and |R4−R5|≦tr and |R6−R5|≦tr, otherwise d4=0
In these expressions tp and tr are predetermined positive threshold values that are chosen to exclude pixel values that are separated from the central pixel (R5) by any edges in the pixel neighborhood shown in
P
M=median[(P1−P5), (P2−P5), (P3−P5), (P4−P5), (P5−P5), (P6−P5), (P7−P5), (P8−P5), (P9−P5)]
R
M=median[(R1−R5), (R2−R5), (R3−R5), (R4−R5), (R5−P5), (R6−R5), (R7−R5), (R8−R5), (R9−R5)]
RM is set equal to PM and, noting that adding or subtracting a constant from all the terms inside the median operator does not change the order of the terms, the expression is solved for R5.
R
5=median(R1, R2, R3, R4, R5, R6, R7, R8, R9)−median(P1, P2, P3, P4, P5, P6, P7, P8, P9)+P5.
An alternate method is to simultaneously use both the first and second spatial pixel differences averaged over a valid pixel neighborhood. Again referring to
R
5
=[c
1(P1−P5+R1)+c2(P2−P5+R2)+c3(P3−P5+R3)+c4(P4−P5+R4)+c5(P5−P5+R5)+c6(P6−P5+R6)+c7(P7−P5+R7)+c8(P8−P5+R8)+c9(P9−P5+R9)]/2(c1+c2+c3+c4+c5+c6+c7+c8+c9)+[d1(−P1+2P5−P9+R1+R9)/2+d2(−P2+2P5−P8+R2+R8)/2+d3(−P3+2P5−P7+R3+R7)/2+d4(−P4+2P5−P6+R4+R6)/2+R5]/2(d1+d2+d3+d4+1)
The weighting coefficients c1 through c9 and d1 through d4 are the same as above. The average of the first and second spatial pixel differences can also be replaced by a weighted average, and the weights may be either fixed or calculated according to, for example, how close R5 is to an edge.
Alternate methods to the four examples discussed above include using the maximum of the first spatial pixel differences, the minimum of the first spatial pixel differences, and an adaptive directional median filter. Still more alternate methods are possible by utilizing other panchromatic characteristics that are known to those skilled in the art in conjunction with other well-known noise reduction methods such as, but not limited to, infinite impulse response (IIR) filtering and singular value decomposition (SVD).
It is also well known by those skilled in the art that pixel neighborhoods such as depicted in
Returning to
In
In
The weighting coefficients c1 through c9 are computed from differences in panchromatic values.
c
1=1 if |P1−P5|≦t, otherwise c1=0
c
2=1 if |P2−P5|≦t, otherwise c2=0
c
3=1 if |P3−P5|≦t, otherwise c3=0
c
4=1 if |P4−P5|≦t, otherwise c4=0
c
5=1 if |P5−P5|≦t, otherwise c5=0
c
6=1 if |P6−P5|≦t, otherwise c6=0
c
7=1 if |P7−P5|≦t, otherwise c7=0
c
8=1 if |P8−P5|≦t, otherwise c8=0
c
9=1 if |P9−P5|≦t, otherwise c9=0
In these expressions t is a predetermined threshold value that is chosen to exclude pixel values that are separated from the central pixel (P5) by any edges in the pixel neighborhood shown in
P
H=median (P4, P5, P6)
P
B=median (P1, P5, P9)
P
V=median (P2, P5, P8)
P
S=median (P3, P5, P7)
The noise-reduced value for P5 corresponds to the panchromatic median value that is closest to original panchromatic value associated with P5.
P
5
=P
H if |PH−P5|≦{|PB−P5|, |PV−P5|, |PS−P5|}
P
5
=P
B if |PB−P5|≦{|PH−P5|, |PV−P5|, |PS−P5|}
P
5
=P
V if |PV−P5|≦{|PH−P5|, |PB−P5|, |PS−P5|}
P
5
=P
S if |PS−P5|≦{|PH−P5|, |PB−P5|, |PV−P5|}
Alternate schemes for employing adaptive median filters are well known in the art and can be used.
In addition to the methods described above, other well-known noise reduction methods such as, but not limited to, infinite impulse response (IIR) filtering and singular value decomposition (SVD) could be used.
It is also well known by those skilled in the art that pixel neighborhoods such as depicted in
The noise reduction algorithms disclosed in the preferred embodiments of the present invention can be employed in a variety of user contexts and environments. Exemplary contexts and environments include, without limitation, wholesale digital photofinishing (which involves exemplary process steps or stages such as film in, digital processing, prints out), retail digital photofinishing (film in, digital processing, prints out), home printing (home scanned film or digital images, digital processing, prints out), desktop software (software that applies algorithms to digital prints to make them better—or even just to change them), digital fulfillment (digital images in—from media or over the web, digital processing, with images out—in digital form on media, digital form over the web, or printed on hard-copy prints), kiosks (digital or scanned input, digital processing, digital or scanned output), mobile devices (e.g., PDA or cell phone that can be used as a processing unit, a display unit, or a unit to give processing instructions), and as a service offered via the World Wide Web.
In each case, the noise reduction algorithms can stand alone or can be a component of a larger system solution. Furthermore, the interfaces with the algorithm, e.g., the scanning or input, the digital processing, the display to a user (if needed), the input of user requests or processing instructions (if needed), the output, can each be on the same or different devices and physical locations, and communication between the devices and locations can be via public or private network connections, or media based communication. Where consistent with the foregoing disclosure of the present invention, the algorithms themselves can be fully automatic, can have user input (be fully or partially manual), can have user or operator review to accept/reject the result, or can be assisted by metadata (metadata that can be user supplied, supplied by a measuring device (e.g. in a camera), or determined by an algorithm). Moreover, the algorithms can interface with a variety of workflow user interface schemes.
The noise reduction algorithms disclosed herein in accordance with the invention can have interior components that utilize various data detection and reduction techniques (e.g., face detection, eye detection, skin detection, flash detection).
The invention has been described in detail with particular reference to certain preferred embodiments thereof, but it will be understood that variations and modifications can be effected within the spirit and scope of the invention.
Reference is made to commonly-assigned U.S. Ser. No. 11/558,571, filed Nov. 10, 2007, of James E. Adams, Jr., et al., entitled “NOISE REDUCTION OF PANCHROMATIC AND COLOR IMAGE”.
