JPH0766442B2 - Image edge enhancement method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of application] The present invention relates to a plate-making scanner (for color, black and white), a facsimile, etc., by photoelectrically scanning an original image to read image data of the original image, The present invention relates to a method of emphasizing a contour part included in.

[Conventional technology]

In enhancing the contour of the duplicate image, a signal of the pixel of interest (hereinafter referred to as a sharp signal S) and an average signal (hereinafter referred to as unsharp signal U) of a region in the vicinity of the pixel of interest (which may include the pixel of interest) are scanned in the original image. , And the original signal for contour enhancement is generated from the difference signal (S−U) between the two. Usually, in order to make the strength of the contour enhancement appropriate, the original signal of the contour enhancement is multiplied by an appropriate coefficient k and amplified to obtain the contour enhancement signal k × (S−U). Is added to the sharp signal S to obtain the image data of the pixel of interest, thereby enhancing the contour of the image.

[Problems to be Solved by the Invention]

However, since the conventional contour enhancement processing is configured as described above, uniform processing is performed on all pixels. Therefore, there is a problem that unnecessary regularity is generated by the contour emphasis processing, and for example, periodic jagged edges (JAG) generated in a straight line portion having a slight angle with the scanning direction are also emphasized.

Further, when the contour emphasis processing is performed in the boundary portion where the density difference is large and the density change is small, for example, as shown in the car pattern shown in FIG. 11, the background black area BR to the gray area of the vehicle body
There is a problem in that false contours PL having different hue densities are generated, such as white stripes appearing slightly in the GR. This false contour PL corresponds to the overshoot portion of the contour-enhanced signal, as shown in FIG. 12 which is an arrow view of the line AA 'in FIG.

[Object of the Invention]

The present invention has been made in consideration of the above circumstances, and suppresses the occurrence of unnecessary regularity due to the contour emphasis process, thereby suppressing the JAG emphasis and the occurrence of false contours having different hue densities. The purpose of the present invention is to obtain an image contour enhancement method that can obtain a visually natural image.

[Means for Solving the Problems]

An image contour emphasizing method according to the present invention is a method for emphasizing contours on image data of an original image obtained by reading an image of an original image for each pixel, and first, specifying each pixel on the original image as a target pixel in sequence. , The sharp signal is extracted from the pixel of interest.

Then, the unsharp signal for each of the target pixels is obtained based on the image data of the region in the vicinity of each of the sequentially specified target pixels.

Then, a first difference signal indicating the difference between the sharp signal and the unsharp signal is obtained, and the first difference signal is limited to the range from 0.5 to 1.5 and corresponds to each of the sequentially specified target pixels. Then, the second difference signal is obtained by multiplying the differently set random number coefficient, the enhanced signal is obtained by linear addition of the sharp signal and the second difference signal, and the enhanced signal is used to enhance the contour of the pixel of interest. The image data is already used.

[Action]

According to the present invention, the second difference signal is obtained by multiplying the first difference signal by different random number coefficients corresponding to the respective target pixels, so that the spatial regularity of the emphasized signal is suppressed. Moreover, the random number coefficient for multiplying the first difference signal is 0.
Since the range is limited to 5 to 1.5, the second difference signal can be softly changed for each pixel while maintaining the original size of the first difference signal to some extent.

〔Example〕

A. Overall Configuration and Schematic Operation FIG. 2 is a schematic block diagram of a plate-making scanner to which an embodiment of the present invention is applied. In the figure, the image of the original image 100 is read for each pixel by the scanning reading device 200,
The image signal thus obtained is transferred to the image processing device 300. The image processing device 300 includes a contour enhancement device 400 having a function described below, and performs a process such as contour enhancement on an input image signal. Then, the processed image signal is given to the scanning recording apparatus 500. Scanning recorder 5
00 converts the image signal into a halftone dot signal, and based on that,
A halftone dot image is exposed and recorded on the photosensitive film 600.

B. Related Proposed Example Prior to explaining the embodiments of the present invention, a related proposed example will be described. FIG. 1A is a flowchart showing an image contour enhancement method which is a proposed example related to the present invention.
This processing is performed by the contour emphasis device 400 shown in FIG.

First, in step S11 of FIG. 1A, a plurality of masking matrices for forming the unsharp signal U are prepared and stored in a predetermined memory area as a preparatory step prior to the emphasis processing.

This masking matrix is, for example, a simple average type that performs uniform weighting W1 as shown in FIG. 3A, or a differential type that performs greater weighting W2 for pixels closer to the center as shown in FIG. 3B. Different types are prepared. Also, as for the range in which the mask operates, those operating in various width ranges are prepared. Actually digital processing
Multiple masks with different weights are prepared as shown in FIG. 4C.

In the next step S12, the sharp signal S corresponding to the pixel of interest is obtained, and one or more masking matrices are sequentially or randomly extracted from the plurality of masking matrices prepared in step S11. Apply. By such processing, different masking matrices are applied to adjacent pixels to form respective unsharp signals U having different characteristics.

In step S13, the sharp signal S obtained in step S12
And the unsharp signal U, the difference signal (S−U) is obtained, and the emphasized signal S + k (S−U) is obtained.

Generally, when a masking matrix is applied to the sharp signal S to obtain the emphasized signal S + k (S−U), the sharpened signal S is obtained as shown in FIG.
Undershoot US and overshoot OS are formed before and after the edge as shown in FIG. 3B. Step S11
If a plurality of masking matrices prepared in step 1 are used, the shapes, heights, and widths of the undershoot US and the overshoot OS are variously changed depending on the difference in the masking matrix, and each pixel has a different shape. Also, in FIGS. 4A to 4C showing the digital processing, it is shown that the heights e1 to e3 of the undershoot US and the overshoot OS are different from each other due to the difference in the masking matrix.

In step S14, the emphasized signal S + obtained in step S13
The contour enhancement processing is completed by setting k (SU) to new image data corresponding to each pixel.

Even if this contour emphasis processing is performed, since overshoots and undershoots have different shapes for adjacent pixels, occurrence of unnecessary regularity due to the emphasis processing is suppressed. This regularity can be sufficiently reduced by increasing the number of masking matrices to be prepared, and it is possible to suppress the occurrence of JAG and the false contour of different hue density.

Next, a circuit configuration for realizing the above processing will be described. FIG. 5 shows the contour enhancement device 400 shown in FIG.
3 is a circuit diagram showing an example of the configuration of FIG.

The signals Y, M, C, and K of each color plate generated for each pixel are passed through a color correction circuit (not shown) or the like, and then the multiplier 4 in the contour enhancement device 400
1Y, 41M, 41C, 41K, given a predetermined coefficient a _y , a
It is multiplied by _m , a _c , a _k . The signals Y, M, C, and K are also given to the delay circuits 51Y, 51M, 51C, and 51M, and the timing adjustment for the processing described later is performed there.

The multiplication results of the multipliers 41Y to 41K are mutually added by the adder 61, and as shown in the following formula (1), the signals Y, M, and
A sharp signal S is generated by the weighted average of C and K.

_{S = a y · Y + a} m · M + a c · C + a k · Ka y + a m + a c + a k = 1 ... (1) sharp signal S is applied to the line buffer memory 71,
It is written according to the address A _p from the address generation circuit 81. The line buffer memory 71 is, as shown in FIG.
The memory capacity (N, which is determined by the number of pixels N scanned in one scan and the number of rows L necessary for masking described later) (N
+ L). For example, when masking 5 × 5, L = 5.

The sharp signal S written as described above is used as a sharp signal S _ij (i, j = 0 to 4) corresponding to the pixels in the matrix of 5 × 5 shown in FIG. 7, for example, in the line buffer memory.
It is read from 71 and the following operations are performed.

Each of the coefficient memories 91 to 96 stores weighting coefficients n _ij (i, j = 0 to 4) of 6 different groups. The coefficient n _ij is for directly generating the difference signal (S−U) from the sharp signal S, and the weighting coefficient m _ij (i, j = 0) for generating the unsharp signal U is used.
~ 4) is given by the following equations (2) and (3).

In this way, as shown in FIG. 8, the coefficients n _ij (i, j = 0
.About.4) and act on each of the sharp signals S _ij (i, j = 0 to 4) shown in FIG. 7 to generate a difference signal (S−U). Address A from address generation circuit 81
For example, _M has a configuration in which one of the coefficient memories 91 to 96 is selected by the upper 3 bits and one of the coefficients n _ij therein is selected by the lower 5 bits. Address by changing the upper three bits of the A _M for each pixel adjacent, it is possible to select each of the coefficient memories 91 to 96 randomly or sequentially. Therefore, a difference signal (S-U) having different characteristics for each pixel is generated.

In FIG. 5, each sharp signal S _ij and coefficient n
Each of _ij is multiplied by the multiplier 42. The 25 multiplication results corresponding to 5 × 5 matrix pixels are cumulatively added in the adder 62.

The difference signal (SU) obtained as described above is added by the adder.
It is given to each of the multipliers 43Y, 43M, 43C, 43K from 62,
The enhancement factors b _y , b _m , b _c , b _k for each color plate are multiplied.

These multiplication results are given to the adders 63Y, 63M, 63C, and 63K, respectively, and are added to the corresponding signals Y, M, C, and K delayed by the delay circuits 51Y to 51K, respectively, and the following colors ( 4a)
The enhanced signals EY, EM, EC, EK as shown in (4b) are generated.

_{EY = Y + b y (S} -U) ... (4a) EM = M + b m (S-U) ... (4b) EC = C + b c (S-U) ... (4c) EK = K + b k (S-U) ... ( 4d) Since the difference signal (S−U) has different characteristics for each pixel as described above, the overshoot portion and undershoot portion of the emphasized signals EY, EM, EC, EK also have different characteristics for adjacent pixels. Have. Therefore, while the contour is emphasized, the occurrence of unnecessary regularity due to the emphasis processing is suppressed, and the occurrence of JAG and the false contour having different hue density are suppressed. Also, an image with a visually natural impression can be obtained.

In addition, although the edge enhancement processing for the 5 × 5 matrix of pixels has been described, the same can be applied to the edge enhancement processing for regions of other sizes. Further, the sizes of the matrixes of the coefficients n _ij prepared in advance may be different from each other, that is, 3 × 3 masking and 5 × 5 unsharp mask may be mixed and used. Further, the number of groups of the coefficient n _ij stored in the coefficient memories 91 to 96 may be plural, and the larger the number, the more sufficient effect can be obtained.

C. Details of One Embodiment Next, details of one embodiment of the present invention in which the above-mentioned proposed example is further improved will be described. FIG. 1B is a flowchart showing a contour emphasizing method which is an embodiment of the present invention. This processing is also performed by the contour emphasizing device 400 shown in FIG.

First, in step S21, a sharp signal S is obtained from the pixel of interest and an unsharp signal U is obtained from a region near the pixel of interest. The unsharp signal U is given by a predetermined masking matrix. Further, in step S22, the sharp signal S
And a first difference signal (S-
U).

In step S23, the first difference signal (S−U) is multiplied by a different coefficient r given to each pixel to obtain a second difference signal r (S
-U). Since the coefficient r changes irregularly or in a fixed order, a sufficiently different second difference signal r (S-U) is generated even for adjacent pixels.

In step S24, the emphasized signal S + k · r (S−U) based on the second difference signal r (S−U) is obtained. Further, in step S25, the edge enhancement processing is completed by associating the enhanced signal S + k · r (S−U) with each pixel. Second
Difference signal r (S−U) has a sufficient difference between adjacent pixels, and the overshoot portion and undershoot portion of the emphasized signal S + k · r (S−U) are sufficient for each adjacent pixel. Since it has different shapes in the shape, unnecessary regularity due to emphasis processing is suppressed, JAG and false contours with different hue density can be suppressed, and an image with a visually natural impression can be suppressed. Is obtained.

Next, a circuit configuration for realizing the above processing will be described. FIG. 9 shows the contour enhancement device 400 shown in FIG.
3 is a circuit diagram showing another example of the configuration of FIG.

In the same manner as the circuit shown in FIG. 5, sharp signal S is applied to the line buffer memory 71, it is written according to the address A _P from the address generating circuit 82.

Sharp signal read from the line buffer memory 71
Each of the S _ij is multiplied by each of the weighting coefficients n _ij (i, j = 0 to 4) read from the coefficient memory 97 in the multiplier 42. The coefficient n _ij is specified by the address A _N from the address generation circuit 82. Further, those 25 multiplication results are cumulatively added in the adder 62 to obtain the first difference signal (S
-U) is generated.

The first difference signal (SU) is provided to the multiplier 44 in the second difference signal generator 52. The multiplier 44 is also supplied with the random number r from the random number memory 98, and generates the second difference signal r (S−U) which is the multiplication result between them. The random number memory 98 is given an address A _R that changes for each pixel from the address generation circuit 82, and generates a random number r that is different for each pixel.

If a numerical value in the range of 0.5 <r <1.5 is stored in the random number memory 98 as the random number r, the second difference signal r (S−U)
Changes in the range of 0.5 (S−U) <r (S−U) <1.5 (S−U), and the pixel size is maintained while maintaining the original magnitude of the first difference signal (S−U). The second difference signal r that changes with each
(SU) is given.

The second difference signal r (SU) is multiplied by the multipliers 43Y-4 shown in FIG.
It is given to 2K, and is then subjected to the same processing as the above-mentioned proposed example. Each enhancement already signals EY~EK generated, enhancement coefficient different for each pixel _{r · b y, r · b} m, r · b c, since given the respective r · b _k, by enhancement processing The generation of unnecessary regularity is suppressed, the generation of JAG and false contours with different hue densities are suppressed, and a visually natural image is obtained.

FIG. 10 is a block diagram showing a configuration of a second difference signal generator configured by a look-up table (LUT). Memory area M1 in LUT44a that constitutes the second difference signal generator 52a
Each of ~ M8 stores a different range of grape values. For example, it is assumed that a numerical value in a relatively large range is stored in the upper memory area M1 with respect to the input random number r, and a numerical value in a relatively small value range is stored toward the lower memory area M8. To do. The numerical value in each memory area M1 to M8 is a constant multiple of the corresponding random number r within each range.

Further, the first difference signal (S−U) is given to the Y decoder DY, and Y for selecting one of the memory areas M1 to M8 is selected.
Acts as an address. When the first difference signal (SU) is relatively large, the upper memory area on the memory area M1 side is selected, and when it is relatively small, the lower memory area on the memory area M8 side is selected. . In this way, a group of numerical values in the range that reflects the magnitude of the first difference signal (S-U) is selected.

The random number r is given to the X decoder DX and functions as an X address for selecting a specific numerical value stored in one of the selected memory areas M1 to M8. The numerical value selected in this way is given to the multipliers 43Y to 43K shown in FIG. 5 in the subsequent stage as the second difference signal R (SU).
The second difference signal R (S-U) reflects the magnitude of the first difference signal (S-U) and varies within the range of each of the memory areas M1 to M8. The generation of unnecessary regularity is suppressed, the generation of JAG and false contours with different hue densities are suppressed, and a visually natural image is obtained.

If the LUT is used, the first difference signal (S−U) and the second difference signal (S−U)
Various non-linear relationships can be applied to the difference signal R (S-U). For example, the first difference signal (S-U)
Becomes a predetermined level or higher, the second difference signal R (S-U)
The LUT may be configured so that is saturated.

〔The invention's effect〕

As described above, according to the present invention, since the second difference signal is obtained by multiplying the first difference signal by different random number coefficients corresponding to the respective target pixels, the spatial regularity of the emphasized signal is suppressed. To be done.

Therefore, by suppressing the occurrence of unnecessary regularity due to the emphasis processing at both edges of the emphasized signal, it is possible to suppress the JAG emphasis and the generation of false contours with different hue densities, and to create a visually natural impression. It is possible to obtain an image contour enhancement method by which an image is obtained.

Moreover, since the random number coefficient for multiplying the first difference signal is limited to the range of 0.5 to 1.5, the second difference signal is pixel by pixel while maintaining the original size of the first difference signal to some extent. It can be changed softly.

[Brief description of drawings]

FIG. 1A is a flow chart of a contour enhancement method according to a proposed example of the present invention, FIG. 1B is a flow chart of a contour enhancement method according to an embodiment of the present invention, and FIG. 2 is a flowchart of a plate making scanner to which the embodiment of the present invention is applied. Block diagrams, FIGS. 3A and 3B are diagrams showing a weighting distribution of a masking matrix, FIGS. 4A to 4C are diagrams showing a digital weighting state, and FIG. 5 is an example of a configuration of a contour emphasis device. FIG. 6 is a circuit diagram showing the configuration of a line buffer memory, FIG. 7 is a diagram showing matrix-shaped sharp signals, FIG. 8 is a diagram showing distribution of weighting coefficients, and FIG. FIG. 10 is a circuit diagram showing another example of the configuration, FIG. 10 is a block diagram showing the configuration of the second difference signal generator, FIG. 11 is a diagram showing a state of conventional contour enhancement, and FIG. 12 is shown in FIG. It is a figure which shows an image signal. 400 …… Contour enhancement device, S, S _ij …… Sharp signal, U …… Unsharp signal, m _ij , n _ij …… Weighting coefficient, 91 to 96 …… Coefficient memory, EY, EM, EC, EK …… Enhanced signal, S-U ... First difference signal, r ... Random number, r (S-U), R (S-U) ... Second difference signal

 ─────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A 61-121589 (JP, A) JP-A 63-155273 (JP, A) JP-A 59-141871 (JP, A)

Claims (1)

[Claims] 1. A method of applying contour enhancement to image data of an original image obtained by reading an image of the original image pixel by pixel, comprising: (a) sequentially specifying each pixel on the original image as a pixel of interest, Extracting a sharp signal from the pixel of interest, (b) obtaining an unsharp signal based on image data of regions in the vicinity of each of the sequentially identified pixels of interest, (c) the sharp signal and the unsharp signal Obtaining a first difference signal indicating a difference from the sharp signal, and (d) corresponding to each of the sequentially specified pixels of interest limited to the range from 0.5 to 1.5 in the first difference signal. Multiplying differently set random number coefficients to obtain a second difference signal; (e) obtaining an enhanced signal by linear addition of the sharp signal and the second difference signal, and obtaining the enhanced signal Attention An image contour enhancement method including: a step of providing image data with pixel contour enhancement.