Image analysis – Image transformation or preprocessing – Changing the image coordinates
Reexamination Certificate
2000-12-29
2002-06-18
Boudreau, Leo (Department: 2621)
Image analysis
Image transformation or preprocessing
Changing the image coordinates
Reexamination Certificate
active
06408108
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image processing apparatus, and more specifically, to a measure to cope with noises generated by an image processing apparatus which carries out image modification processing to invert at least a portion of a binary signal obtained through binary conversion of an image signal on which resolution correction for improving spatial resolution has been carried out.
2. Description of the Related Art
An image inputting apparatus has been known which obtains an image signal carrying image information of an original by scanning the original with an image sensor or the like and converting the original image into an electric signal (for example, a scanner, or a stencil making apparatus). To improve optical resolution, it is general for such an image inputting apparatus to carry out resolution correction for improving the spatial resolution of an image (for example, MTF correction) on an image signal having been obtained.
After an image signal have been binary converted, if the image wherein white pixels are scarcely seen in a black pixel dominant area (a black pixel area) is printed by a printer for example, the white pixels in the black pixel area are not so conspicuous. However, when such an image is printed by a printer after inversion processing or inside whitening processing has been carried out thereon, black pixels which have been inverted from scarce white pixels in a white pixel dominant area (a white pixel area) which has been inverted from the black pixel area become conspicuous. This visual phenomenon has been known.
FIG. 10
is a diagram showing this visual phenomenon and shows a relation between the human visual sense and a ratio of black pixels within an area printed by a printer. In the area where black pixels are dominant (a portion to the right in the figure), the ability to identify a predetermined number of white pixels (A) therein is low and the white pixels are not conspicuous. In the area (a portion to the left of
FIG. 10
) where black pixels are scarce, that is, the area wherein white pixels are dominant, the ability to identify the same number of black pixels (B) therein is high and the black pixels are very conspicuous.
Especially in stencil making printing for example, the finish itself of printing tends to extend areas of black pixels due to an effect so-called smear of ink. Therefore, black pixels in a white pixel dominant area become much more conspicuous.
When MTF correction is carried out on an image signal, the above phenomenon becomes more conspicuous. This is because that a white noise signal mixed with an image signal output from an image sensor or the like and having the density of black becomes too enhanced by binary conversion to be at the level of white after MTF correction has been carried out on the image. Therefore, in some cases, after binary conversion has been carried out on a signal having been MTF corrected and a predetermined area has been specified by a digitizer or the like, if signal inverting processing (hereinafter called image modification processing), for example inversion processing and inside whitening processing on the signal in the predetermined area or hatching processing on the area whose inside has been whitened, is carried out on at least a portion of the signal having been binary converted, a portion of the image in the predetermined area which should be inverted from black to white is not inverted and remains black. As a result, black noises in white background become conspicuous.
Hereinafter, the reason why this phenomenon becomes conspicuous after binary conversion has been carried out on a signal having been MTF corrected. MTF correction will first be explained briefly.
FIG. 11
shows a general MTF correction method. In this MTF correction, the value of a target pixel A to be processed is doubled (D=A×2). A value E which is a sum of the values B and C of pixels neighboring with A in the horizontal (main scanning) direction (E=B+C) is divided by 2 to make F (F=E÷2), and F is subtracted from D to make G (G=D−F). In this manner, when the value A of the target pixel is different from the values of neighboring pixels B and C, the value A is changed to a larger one if A is larger than those, and to a smaller one if otherwise. Therefore, by sequentially applying this MTF correction along the main scanning direction, optical resolution along the main scanning direction can be improved maintaining the average (the average density) of the values in the image signal. It is needless to say that vertical optical resolution can be improved as well by applying this correction to pixels neighboring with the target pixel in vertical direction.
The phenomenon in which black pixels due to a noise signal are created by inversion processing on a signal having been MTF corrected will be explained next.
FIG. 12
shows an example of inversion processing. The basis of the inversion processing is to carry out logical inversion on a target pixel A before processing. In the block diagram and logic expression in
FIG. 12
, a circuit for selecting whether to carry out inversion processing is further added to the above basis.
By carrying out this inversion processing, processing shown at the bottom of
FIG. 12
corresponding to a selecting signal C from the selecting circuit is carried out on the target pixel A before inversion. In the case of non-inversion (the selecting signal C is 0), a target pixel D after the processing has the same logic as the target pixel A before the processing, while in the case of inversion (the selecting signal C is 1), the target pixel D after the processing has the logic inverted from that of the target pixel A before the processing. The case where this inversion processing is actually applied to an image signal will be explained next.
An image signal carrying image information of an original is obtained by scanning the original by using an image sensor and converting the original image into an electric signal, and the image signal is converted from analogue to digital by using an A/D converter. An example of the values of the digital signal is shown as a table in
FIG. 13
(A). In
FIG. 13
(A), the value 60 of a pixel K
3
is a little smaller than the value 80 of the surrounding pixels, and shows a small noise component. However, when binary conversion is carried out on these pixels wherein the values 50 or smaller are converted to white pixels, both the pixel K
3
and the surrounding pixels are binary converted to black pixels. Therefore, the pixel K
3
does not turn out to be a noise which can be recognized as a white pixel in a black pixel area. Consequently, when the inversion processing is carried out thereon, the pixel K
3
becomes a white pixel and it does not turn out to be a noise recognized as a black pixel. However, when the inversion processing is carried out after the MTF correction on the pixels, the pixel K
3
becomes a black noise pixel.
The values after MTF correction carried out on the signal shown in
FIG. 13
(A) are shown as a table in
FIG. 13
(B). As shown in
FIG. 13
(B), the value of the pixel K
3
has changed from 60 to 40 after the MTF correction, and the effect as a noise is greater than before the correction.
FIG. 14
(A) is a table showing data after binary conversion on the signal in
FIG. 13
(B) wherein the pixels having the values 50 or smaller have been converted into white pixels and the pixels having the values larger than 50 have been converted into black pixels. The noise component in the pixel K
3
is recognized as a white pixel in an area of black pixels.
FIG. 14
(B) shows a table after inversion processing on the pixels in
FIG. 14
(A), and
FIG. 14
(C) shows a visual expression of the pixels in
FIG. 14
(B). As shown in
FIGS. 14
(B) and
14
(C), the pixel K
3
becomes a white pixel if no MTF correction has been carried out thereon, while it becomes a black pixel after the MTF correction. As a result, this black pixel become
Hara Yoshikazu
Murayama Yoshikazu
Nishihata Masahiro
Boudreau Leo
Nixon & Peabody LLP
Patel Kanji
Riso Kagaku Corporation
Studebaker Donald R.
LandOfFree
Image processing apparatus 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 apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2980866