Image analysis – Color image processing
Reexamination Certificate
2000-07-27
2004-04-20
Mehta, Bhavesh M. (Department: 2625)
Image analysis
Color image processing
C358S514000, C358S525000, C348S275000
Reexamination Certificate
active
06724932
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image processing method and an image processor for estimating a signal value at all pixel positions by employing image data which does not have a signal value representing all colors at all pixel positions, like image data representing a color image obtained by an imaging device such as a single plate charge-coupled device (CCD), and to a computer readable storage medium storing a program for causing a computer to execute the image processing method.
2. Description of the Related Art
As an imaging device such as a CCD employed in a digital camera, an imaging device in which a plurality of kinds of photoelectric conversion elements differing in spectral sensitivity are alternately arrayed on the same surface (hereinafter referred to a single plate CCD) is known. In the case of the single plate CCD, in which photoelectric conversion elements having spectral sensitivities for red (R), green (G), and blue (B), i.e., photoelectric conversion elements for an R-channel, a G-channel, and a B-channel are alternately arrayed, a set of three continuous photoelectric conversion elements for an R-channel, a G-channel, and a B-channel constitutes a single pixel. In such a single plate CCD, however, shift of color alignment or a false color sometimes occurs because the R-signal, G-signal, and B-signal values at each pixel cannot be obtained at the same pixel position. Also, since the number of photoelectric conversion elements for channels is less than the total number of pixels constituting a single plate CCD, an image with high resolution cannot be obtained. For instance, in a single plate CCD where photoelectric conversion elements for an R-channel, a G-channel, and a B-channel are alternately arrayed, the number of photoelectric conversion elements for channels is only one-third the total number of elements and therefore the resolution becomes ⅓ compared with a monochrome imaging device having the same number of elements. Because of this, a method has been proposed in which a signal value is computed by an interpolating process at a position where photoelectric conversion elements for an R-channel, a G-channel, and a B-channel are not present. However, there are cases where only an interpolating process causes a false color to occur at a position where a signal value changes considerably. In this case, although the occurrence of a false color can be prevented by using an optical low-pass filter in an imaging system, or by performing a smoothing process on an image signal through a low-pass filter, there is a problem that the resolution will degrade.
Here, human visual sense characteristics are higher in sensitivity with respect to brightness than with respect to color. For this reason, a method has been proposed in which a high-frequency brightness signal (which represents the brightness of each pixel) and a low-frequency color signal (obtained by the above-mentioned interpolating process and the smoothing process which employs a low-pass filter) are generated from a color image signal obtained by a single plate CCD and color image signals are reconstituted by the generated brightness signal and color signal (Japanese Unexamined Patent Publication No. 10(1998)-200906). This method can obtain color image signals capable of reproducing an image whose resolution is apparently high, because much information about a brightness component whose sensitivity is high in the human visual sense characteristic can be given.
Incidentally, a CCD with a honeycomb array of pixels arrayed checkerwise as shown in
FIG. 16
, for example, is known as a single plate CCD (e.g., Japanese Unexamined Patent Publication No. 10(1998)-136391). This is also called a checkered pixel array. There is known another CCD that has a Bayer array of pixels arrayed in square form, as shown in FIG.
17
. This is also called a square pixel array. The single plate CCD having such an array of pixels also has the problem of false color. Furthermore, a method is known which is based on the assumption that the ratio of an r-signal, a g-signal, and a b-signal is approximately constant at local regions of an image (Japanese Unexamined Patent Publication No. 9(1997)-214989). In order to remove a false color from a light-quantity based signal obtained by the aforementioned CCD having a Bayer array of pixels, in the vertical or horizontal direction of the CCD a g-signal is multiplied by the ratio of a r-signal and a g-signal obtained at the adjacent line, whereby an r-signal at that line is computed. More specifically, in an array of pixels shown in
FIG. 18
, in order to compute an r-signal r
12
at a g
12
-pixel position, a g
11
-signal at an r
11
-pixel position is first computed by an equation of (g
6
+g
16
)/2. Based on an assumption of r
11
:g
11
=r
12
:g
12
, the r
12
-signal at the g
12
-pixel position is computed by an equation of r
12
=g
12
×r
11
/g
11
.
However, in the method disclosed in the above-mentioned Japanese Unexamined Patent Publication No. 10(1998)-200906), etc., even if a smoothing process through a low-pass filter is performed on an image signal obtained by the single plate CCD, a high-frequency component in the actual image has been folded back into the image. Because of this, moire due to aliasing noise cannot be removed, and consequently, a false color cannot be sufficiently removed.
On the other hand, the method disclosed in the aforementioned Japanese Unexamined Patent Publication No. 9(1997)-214989 is capable of removing a false color effectively. Particularly, this method is based on the assumption that the light-quantity ratio of r:g:b is constant at local regions of an image, and in the case of an analog signal in which the ratio of an R-signal, a G-signal, and a B-signal obtained is proportional to a quantity of light, a false signal in an image signal obtained by a CCD having a Bayer array of pixels can be effectively removed. However, an image signal obtained by a digital camera does not become r:g:b=R:G:B, because, when light quantities r, g, b are converted to a digital R-signal, G-signal, and B-signal, a signal value is represented by an exponential or logarithmic value such as R=r
0.45
and R=log(r), in order to reduce quantum errors and input a signal to a video circuit in a computer system. Because of this, the method disclosed in the aforementioned Publication No. 9(1997)-214989 can remove a false color from an analog signal whose signal value is proportional to a quantity of light, but cannot remove a false color in the case where a signal value is represented by an exponential or logarithmic value of a quantity of light. In addition, a false color occurs not only in a single plate CCD having a Bayer array of pixels but also in a single plate CCD having a honeycomb array of pixels.
SUMMARY OF THE INVENTION
The present invention has been made in view of the drawbacks found in the aforementioned methods. Accordingly, it is an object of the invention to provide an image processing method and an image processor which are capable of reducing the occurrence of a false color even if a signal is of any type. Another object of the invention is to provide a computer readable storage medium in which a program for causing a computer to execute the image processing method is stored.
For example, in the case where each signal value, which constitutes image data obtained by an imaging device such as a single plate CCD, is represented by the exponential value or logarithmic value of a light quantity, the image processing method according to the present invention has been made based on the assumption that the differences between signal values become equal at local regions of an image.
In accordance with the present invention, there is provided a method of processing image data,
the data representing an image that includes a first pixel, a second pixel, and a third pixel that have a first signal value, a second signal value, and a third signal value that have different sp
Birch & Stewart Kolasch & Birch, LLP
Fuji Photo Film Co. , Ltd.
Mehta Bhavesh M.
Sukhaphadhana Christopher
LandOfFree
Image processing method, image processor, and storage medium does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image processing method, image processor, and storage medium, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing method, image processor, and storage medium will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3223714