JPS58182957A - Sharpness emphasizing system for picture - Google Patents

Abstract

PURPOSE:To emphasize the contour part of a picture and to obtain a picture which is harmonious on the whole by multiplying the difference between a sharp and an unsharp signal by a sharpness emphasis coefficient when the difference is greater than a threshold value, and adding the product to a picture signal. CONSTITUTION:The picture sharp signal S and unsharp signal U are inputted to a subtracter 1, whose subtraction result is inputted to a blind zone circuit 2 whose blind zine is set by threshold values C1 and C2, inputting its output (f) to a multiplying and integrating circuit 3. This circuit 3 inputs the sharpness emphasis coefficient Ki and a picture signal Xi to output a signal Xi' having sharpness emphasized. The circuit 3 after multiplying the output (f) by the coefficient Ki adds the product to the signal Xi to output the addition result as a picture emphasized signal Xi'. Therefore, the sharpness is not emphasized unless the difference signal S-U is greater than the threshold value, and the contour part of the picture is emphasized to obtain the picture which is harmonious on the whole.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an image sharpening and enhancement method used for scanning images such as color scanners for plate making, laser cutter printers, facsimile machines, and televisions.

As a sharpness enhancement method of iii*, a so-called unsharp mask method is known in which a signal obtained by strongly amplifying the difference between a sharp signal and an unsharp signal of an image is added to the original image signal. In other words, set the sharpness signal of the image to 8. Assuming that the unsharp signal is U, the original image signal is X, and K is a coefficient that determines the degree of sharpness enhancement, the emphasized signal Xe is X, , X+K (8-U)
......(1). Here, the unit of signal is concentration. In this conventional sharpness enhancement method,
It emphasizes the contrast not only in areas where the density changes, such as contours with large contrast in the image (but also in areas with relatively small contrast such as film grain, roughness, scratches, and unevenness in the skin of a person).・There is a drawback that it has a negative effect on image quality.

Figure 1 schematically explains this situation, and each song *(S, U+s-U+Xe) in the figure is plotted with time (
(corresponding to position), signal in vertical direction (corresponding to #1 degree)
indicates the size of A difference signal 8 is generated from the sharp signal S and the unsharp signal U obtained by photoelectrically scanning the edge part of the image.
−U is obtained, and this difference signal S−U is amplified to obtain an image signal X(
For example, the sharp signal S may be used), so the final image strength V4 signal X
e emphasizes not only the edge part of the ms (a in the figure) but also the graininess of the parts that should be flat (s, b' in the figure), which has a negative effect on the image quality.

Note that the shape of the sharp signal S, such as film grain or the roughness of a person's skin, can be seen as is in the signal waveform, but the unsharp signal U is blurred, so its signal waveform is smoother than the sharp signal S. .

Therefore, it is an object of the present invention to emphasize the sharpness of an image without emphasizing areas that have a negative impact on image quality, such as film grains and the rough texture of a person's skin. The object of the present invention is to provide a method for sharpening an image to obtain a harmonious image.

This invention will be explained below.

As shown in FIG.
The difference signal 8-U between the unsharp signal U and the unsharp signal U is set to thresholds C1 and C
2, only the portion exceeding the threshold is amplified, and the amplification coefficient is multiplied by the sharpness emphasized image aK to obtain the image signal X.
By adding K, an image enhancement signal X' is obtained.

1iir my signal dog ('-t*2-s, insert B, G, R color separation signals) k corresponding image intensity vI4 signal
/. As the sharp signal 8, one of the image signals XI may be used. Further, the unsharp signal U includes information from a wider image portion than the sharp signal S. That is, FIG. 2 shows an example of a printing path for implementing the present invention, in which the sharp signal 8 and unsharp signal U of the image are input to the subtracter 1, and the subtraction result S-U or the threshold value C1 and Dead band circuit 2 with non-sn+ni set in C1
The output f is input to the multiplication/accumulation circuit 3. Then, in this multiplication/accumulation circuit M3, the sharpness emphasis coefficient is inputted, and one image part number X. Here, the dead band circuit 2 is connected to the circuit 1! as shown in FIG. Between the threshold values C1 and C, (C1>o, C2<0), the output f is O, above C1, f is 5-U-C1, and below C3, f is output. , 5-U-C3. That is, . Further, the multiplication/accumulation circuit 3 multiplies the output f and the sharpness emphasis coefficient by
1 and outputs the addition result as a ghost emphasis signal X1'. That is, x, '-x, 10, -f · Interval (3).

From the above, the difference S-U between the sharp signal 8 and the unsharp signal U is multiplied and
The image enhancement signal Xm' outputted from the accumulation circuit 3 becomes K as follows.

Togi of C8≦5-IJ≦01 x, /=x.

According to such a configuration, the difference signal s as shown in FIG.
If −U is a value of the function of two thresholds C1 and C, then
No. 15 f is O as shown in the figure, so lI! No. 11 is not emphasized in any way. In other words, sharpness enhancement is not performed until the difference signal S-U exceeds the threshold value, and the outlines of the image are emphasized without emphasizing the roughness of the skin of the film grain 1 person, and the overall harmony is maintained. It is possible to obtain a clear image. Also, the threshold C
By appropriately adjusting 1, C2, and the sharpness emphasis coefficient 1, the letter of the image with arbitrary sharpness characteristics can be made 53 times, and the selection according to the type of image 1mK can be made/obtained.

Next, a more detailed circuit example of FIG. 2 will be shown and explained in FIG.
A case where C1m-c,xc is the relationship will be explained. Therefore, the output signal f shown in FIG. 2 is expressed by the following equation.

The sharp signal 8 (n bits) of the image is input to the adder 5, and the unsharp signal U (n bits) is input to the adder 5, passing through the invar string string 4. In this case, the unsharp signal U is set to 10-1 by the inverter array 4, and 1 is added to the carry input CI of the adder 5. The sign bit of 8-U is the sharp signal S
and -0-1 and the carry output Co (CRI) of the adder 5 are added. Here, the sign bit of the sharp signal S is 0, -
Since the sign bit of U-1 is l, the inverted carry output CRI becomes the sign bit of S-U. Therefore, S-U is a negative number when the carry output CRI is -00, and is a positive number or O when CRI≧1.

The threshold value C (n bits) is stored in the register 6, and is input to the adder 8 via the extrinsic OR circuit string 7, and when 8-U or negative, that is, (,')tl -00, the excl- The Shiv-OR circuit array 7 passes the threshold value C up to 1 and adds t! 8 calculates 8-U+C. Also, S-
When U is positive or 0, that is, CRI-1, extal-
The Shiv-OR circuit array 7 inverts the threshold value C, and l is added to the carry input C1 of the adder 8, so the adder 8
- Calculate U-C. In either case, a positive number and a negative number are input to the adder 8, so the carry CR2 of the adder 8=
When it is 0, it is negative, and when it is CR2-1, it is positive or 0.

Note that the carry output CR2 is inverted by the inverter 11 and becomes a sign bit (1 bit).

When the above calculations are completed, the carry output (CRI
, C) are (0,0), (0,1).

(1,0), (1,1), 8-U is 8-
U(-C,-C50-U(0,0≦l5-U(C,S-
It can be seen that the range is U≧C. Also, exclusive
Since the carry output CR2 and the carry output CRI inverted by the inverter 9 are input to the exclusive OR circuit 10, when the output of the exclusive OR circuit WIr10 is 1, 8-U (-C or 8- When U≧C,O, -
〇≦8-U (C. Then, the output of the adder 8 (n bits) is ANDed with the sign bit from the inverter 11 and the output of the exclusive OR circuit 1o in the AND circuit array 12. When -C50-U≦C, it is always 0. Furthermore, when the sharp signal S and the unsharp signal U are each represented by n bits, a sign bit is added to the output f of the AND circuit string ( n+1) bits.In addition, when setting the threshold values C1 and C2 individually, the register 6 and the exclusive OR circuit string 7
It is sufficient to indicate the 0 part as the fifth factor, that is, the threshold value C1 is set in the register 61, and the register 5zKc1 (Maro-C
.

Note that although the sharpness enforcement using the digital circuit has been explained in FIGS. 41 and 5, the same processing can also be performed using an analog circuit. It is also possible to change the sharpness emphasis coefficient depending on the sign of the difference between the sharp signal and the non-sharp signal.

[Brief explanation of drawings]

FIG. 1 is a diagram schematically illustrating the conventional sharpness enhancement method and the sharpness enhancement method of the present invention, FIG. 2 is a block diagram showing an example of a circuit for realizing the present invention method, and FIG. The figure shows an example of the characteristics of the dead band circuit used in this invention.
FIGS. 4 and 5 are circuit diagrams showing specific circuit examples for realizing the present invention. 1...Subtractor, 2...Dead band circuit, 3...Multiply
Cumulative 6...Register, 7...Exclusive OR circuit string.

Claims (1)

[Claims] The feature is that an image enhancement signal is obtained by subtracting an unsharp-1 signal from an image sharp signal, and when the subtraction value is larger than a threshold value, the subtraction value is multiplied by a sharpness enhancement coefficient and added to the image signal. Image sharpness enhancement method.