JP2000115526A - Picture processor and edge processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress blurs in an edge area after high resolution conversion without varying color arrangement provided for an original picture. SOLUTION: This processor is formed of an emphasizing coefficient deciding part 3 obtaining the coefficient of a filter for emphasizing an edge, an edge emphasized picture generation part 4 edge-emphasizing low resolution information obtained from a picture inputting terminal 1 by using the decided emphasis coefficient, an interpolated pixel generating part 5 generating an interpolated pixel by interpolating a pixel value at a designated position from an edge- emphasized low resolution picture, a comparing part 6 obtaining maximum/ minimum density values from the peripheral pixels of the noted pixels of low resolution information to be used for interpolation and comparing these maximum/minimum density values and the interpolated pixel obtained by the part 5 with each other, and a weighting combining part 7 executing weighting to each pixel so as to replace the density value of a pixel position with a density value larger than a maximum value with the maximum value and the density value of a pixel position with a density value smaller than the minimum value with the minimum value from the output of the part 6.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image output device such as a printer which enlarges and resizes input image information and outputs the converted image information, and converts resolution from low-resolution information to high-resolution information by communication between models having different resolutions. The present invention relates to an image processing device that performs similarity between input image information and a template image having a different resolution, and an image processing device that sharpens a blurred image.

[0002]

2. Description of the Related Art Conventionally, various methods have been proposed as a method of generating an interpolation pixel for converting resolution of inputted low-resolution information into high-resolution information. Typically, the conventional interpolation method is a nearest neighbor method in which the same pixel values are arranged closest to the interpolation point, or a bilinear method in which the interpolation pixel value is determined by linearly interpolating the distance between four points surrounding the interpolation point. , The sampling function (f
A cubic convolution method determined using (x) = sinc (x)) is generally used.
These methods are described in the paper "A Study on Interpolation Filters for Image Signal Geometric Transformation and Image Quality" (Enami, et al .: Transactions of the Institute of Electronics, Information and Communication Engineers '86 /11Vol.J69-D No.11 pp1617-1623) and " Recommendation of Image Processing Smoothing and Interpolation ”(Yoichi Miyake: Photographic Industry '87 /9vol45No.9 pp109-113). Hereinafter, these three techniques will be described.

FIG. 9 shows the principle of the nearest neighbor method. Here, the given original image is represented by P, the pixel position in the main scanning direction is represented by i, and the pixel position in the sub-scanning direction is represented by j. That is, P _{i, j} represents a value indicating the luminance or density of the original image at the pixel position (i, j).
Also, the coordinates of the point to be interpolated are (i + Δi, j + Δj)
And Δi and Δj are respectively 0 ≦ Δi <1, 0 ≦ Δ
Any real number given by j <1. As shown in FIG. 9, the nearest neighbor method uses image data for obtaining data of an original image closest to a point to be interpolated, and is expressed by the following equation.

P _{i + Di, j + Dj} = P _{m, n} where Di and Dj mean Δi and Δj, respectively, and m =
[I + Δi + 0.5] and n = [j + Δj + 0.5]. Here, [] represents a Gaussian symbol and indicates that it is the largest integer not exceeding its value. In the case of FIG. 9,
Since (i, j) is the closest pixel _{, Pi + Di, j + Dj} = _{Pi , j} . The shaded area in the figure indicates a pixel range that has the same value as P _{i, j} as in P _{i + Di, j + Dj} .

FIG. 12 shows an example in which an interpolated pixel is generated using the nearest neighbor method. Here, it is shown that the number of pixels is increased from the base 3 × 3 pixels to 9 × 9 pixels. As shown in the figure, each pixel is expanded to 3 × 3 pixels with the same density value for each base pixel.

FIG. 10 shows the principle of the bilinear method. In the bilinear method, image data to be obtained is obtained using four points of original image data surrounding the point (i + Δi, j + Δj) to be interpolated. The operation expression can be represented by the following expression.

[0007]

(Equation 1)

In the bilinear method, the point to be interpolated and the surrounding four
This is a weighted average using the distance to the point as a weight, and the output image is a smoothed version of the original image.

FIG. 11 shows the principle of the cubic convolution method. The image data to be obtained is obtained using 16 original image data surrounding the point (i + Δi, j + Δj) to be interpolated. The operation expression can be represented by the following expression.

[0010]

(Equation 2)

Where f (t) = sinc (t) = sin (πt) / (πt)） 1-2 | t | ² + | t | ³ (0 ≦ | t | <1), 4- 8 | t | +5 | t | ² − | t | ³ (1 ≦ | t | <2), 0 (2 ≦ | t |) At this time, x1 = 1 + Δi, y1 = 1 + Δj, x2 = Δi , Y2 = Δj, x3 = 1−Δi, y3 = 1−Δj, x4 = 2-Δi, y4 = 2-Δj.

FIG. 13 shows an example in which an interpolated pixel is generated using the bilinear method or the cubic convolution method. Here, it is shown that the number of pixels is increased from the base 3 × 3 pixels to 9 × 9 pixels. Since the generated interpolated pixel is obtained from the original pixel around the pixel position, an interpolated pixel whose density value has changed more naturally than in the nearest neighbor method can be obtained, and a more natural image can be obtained.

However, these systems have the following disadvantages. That is, although the nearest neighbor method has an advantage that the configuration is simple, as shown in FIG. 12, since the pixel value is determined uniformly for each block to be enlarged, the blocks are visually conspicuous and the image quality is high. Inferiorly. The bilinear method and the cubic convolution method are generally used for expanding natural pixels. However, as shown in FIG. The image quality is blurred.

In order to solve these drawbacks, a common technique is to use a high-frequency emphasizing filter as shown in FIG. 14 to perform edge emphasis processing on pixels of a target image in advance, and to apply the above-described image processing to the image. A method of obtaining high-resolution information by performing pixel interpolation by the above method, or a method of obtaining sharp image quality by performing edge enhancement processing on high-resolution information obtained by performing pixel interpolation by the above method.

As a conventional example in which the former method is used as preprocessing for resolution conversion, there is a resolution conversion processing method described in JP-A-6-309452. In this method, using the interpolation pixel obtained by the resolution conversion process described above for the low resolution information and the interpolation pixel obtained by the resolution conversion process described above for the edge-enhanced image of the same low resolution information, This method obtains high-resolution information by determining the addition distribution of the density values of the two types of interpolated pixels described above according to the weight information obtained by edge-determining the input low-resolution information for each pixel position. . In this method, when the pixel is regarded as an edge, it acts to strongly weight the interpolated pixel information obtained based on the edge emphasized image.

FIG. 15 shows an example in which an edge enhancement process is performed in advance, and an interpolated pixel is generated when pixel interpolation is performed on the image by the bilinear method. (A) in the figure is an image showing a part of the edge of the input low-resolution information. An image in which edge enhancement has been performed on this pixel is shown as (b), and an image obtained by performing a three-dimensional resolution conversion on each pixel is shown as (c).

[0017]

As shown in FIG.
The image of (c) obtained based on the image of (b) has a clearer edge than the image of FIG. 13 without edge enhancement, but has a darker edge than the original image (a). There are pixels having a color scheme.

As can be seen, in the above conventional example,
By performing the edge enhancement processing, the color scheme of the original image as input data is changed. That is, an area where the density value of the pixel is high changes to a higher value, and an area where the density value of the pixel is low changes to a lower value. Therefore, when resolution conversion is performed on the edge-enhanced image by the interpolation method described above, the original image not subjected to edge enhancement is sharper than the resolution-converted image, but the pixels around the edge are sharper. Has a different color scheme for the input data, and there is a problem that the impression of the image received after the resolution conversion is different from the original image.

Further, in the above-described conventional example, since the edge enhancement processing to be performed is constant irrespective of the number of pixels to be interpolated, the effect of edge enhancement becomes weaker as the number of interpolated pixels becomes larger, such as when the enlargement ratio is large. There is a problem that the obtained high-resolution information is also blurred.

By paying attention to this point, the present invention does not change the color arrangement or shading of a color or monochrome multi-valued original image and suppresses blurring of an edge area after high resolution conversion. In addition, even when a blurred image that is out of focus is given as an input image, it is possible to generate an image in which edges are emphasized and blurring that does not change the color scheme or shading of the input original image is eliminated. It is an object to provide an image processing apparatus, an edge processing method, and a system using the same.

[0021]

According to the present invention, there is provided an image processing apparatus for converting a multi-valued image into an input image and interpolating a pixel value at an arbitrary position on the input image to perform resolution conversion. An image input means for inputting the input image information, an edge-enhanced image generating means for edge-enhancing the input image information, and interpolating and interpolating a pixel value at a designated position on the edge-enhanced input image An interpolated pixel generating means for generating a pixel, and a threshold value is determined by a predetermined method from pixel values of peripheral pixels at a designated position on the input image, and the interpolated pixel generating means generates the threshold value. Weighting means for comparing the pixel value of the interpolated pixel thus obtained, and weighting the interpolated pixel value in accordance with the result of the comparison, thereby synthesizing higher resolution output image information. Has a forming means, and an image output means for outputting the output image information synthesized by said weighting and combining means to the outside.

Here, based on information inputted from the image input means and the information inputted from the information input means for inputting information on the pixel value interpolation processing, a coefficient of a filter used for edge enhancement processing is obtained. An emphasis coefficient determining unit may be further provided. Further, the emphasis coefficient determining means changes the filter coefficient in accordance with a pixel value of a peripheral pixel at a designated position on the input image or in proportion to the number of interpolation pixels input from the information input means. It is good also as a structure to make it.

Alternatively, the comparing means obtains a maximum and a minimum density value as a threshold value in a group of peripheral pixels at a designated position on the input image, and calculates the threshold value and the generated interpolation pixel. By comparison, a weighting coefficient for resetting each pixel value in the generated group of interpolated pixels is determined so as not to exceed the maximum and minimum density values, and the weighting synthesis unit determines the weighting coefficient. The output image information may be combined according to a weighting coefficient.

Alternatively, the comparison means may be configured so that the range of peripheral pixels used for determining the threshold value can be set arbitrarily.

According to another aspect of the present invention, there is provided an image processing apparatus for processing a multivalued image as an input image and performing edge processing on the input image, comprising: an image input unit for inputting the input image information; An edge-enhanced image generating means for generating an image in which the input image information is edge-enhanced, and a threshold value obtained by a predetermined method from a pixel value of a peripheral pixel at a designated position on the input image, A comparison unit that compares the pixel value of the image generated by the edge-enhanced image generation unit with the pixel value of the edge-enhanced image. The image processing apparatus includes a weighting synthesis unit that synthesizes information, and an image output unit that externally outputs the output image information synthesized by the weighting synthesis unit.

According to another aspect of the present invention, there is provided an image processing apparatus for generating information indicating an output image by performing at least edge processing on an input multi-valued image. First processing means for performing image processing including at least edge enhancement processing on the input image, and predetermined using a pixel value in a pixel group including at least a pixel to be subjected to the edge enhancement processing in the input image Comparing means for obtaining a threshold value by the method described above, and comparing the threshold value with the pixel value of the pixel group subjected to the edge enhancement processing in the image output from the first image processing means; Resets the pixel value of the edge-enhanced pixel in the image output from the first image processing means, and outputs the pixel value according to the reset pixel value. And a second processing means for synthesizing an image.

According to another aspect of the present invention, in the edge processing method for a multi-valued image, image processing including at least edge enhancement processing is performed on the multi-valued image, and the edge processing is performed in the input image. A feature amount is obtained for a pixel value of a peripheral pixel group including at least a pixel to be enhanced, and a pixel value of a pixel existing in the edge-enhanced region in the image-processed image according to the feature amount To reset.

[0028]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Some embodiments according to the present invention will be described below with reference to the drawings.

(Embodiment 1) FIG. 1 shows a schematic block diagram of an embodiment of an image processing apparatus for performing resolution conversion according to the present invention. The device according to the present embodiment has a configuration indicated by reference numerals 1 to 9 in the drawings described below.

In FIG. 1, reference numeral 1 denotes an image input terminal of original image information to be subjected to resolution conversion, and generally, a scanner, a digital camera, image data displayed on a screen, and the like are input target data. . The object of the image processing according to the present invention is a color or monochrome multivalued image. If the input image is a color image, the color information is RGB
The color information is not limited to the space information, and various color space information such as YUV space information having brightness and hue as color information can be used.

Reference numeral 2 denotes an information input terminal to which information on interpolation processing relating to resolution conversion, such as the number of pixels to be interpolated, is given. When enlarging an image, the input information is equivalent to the magnification.

Reference numeral 3 denotes an emphasis coefficient determination unit that obtains a filter coefficient for performing edge emphasis based on the number of interpolation pixels input from the input terminal 2. Reference numeral 4 denotes an edge-enhanced image generation unit for edge-enhancing low-resolution information obtained from the input terminal 1 using the determined enhancement coefficient. Reference numeral 5 denotes an interpolated pixel generation unit that generates an interpolated pixel by interpolating a pixel value at a specified position from the edge-enhanced low-resolution image.

Reference numeral 6 compares the pixel of interest on the low-resolution image input from the image input terminal 1 with the output value of the interpolation pixel generation unit 5, and according to the result, a value indicating the degree of weighting of each image value. This is a comparison unit for outputting.

Reference numeral 9 denotes an information input terminal for giving information to the comparing section 6 to arbitrarily designate a peripheral pixel range for detecting the maximum density value and the minimum density value.

A weighting unit 7 weights the pixel of interest on the low-resolution image input from the image input terminal 1 and the output value of the interpolated pixel generation unit 5 in accordance with the weighting output from the comparison unit 6 to perform synthesis. It is a weighting synthesis unit. A method of determining the degree of weighting and a combining method will be described later.

Reference numeral 8 denotes an image output terminal for outputting the high-resolution image information subjected to resolution conversion by the apparatus according to the present embodiment to the outside. The output image generated by the weighting / synthesizing unit 7 is output to the outside.

FIG. 2 shows an example of a system configuration in which the image processing apparatus according to the present invention is employed. This system is composed of devices 21 to 27 described below.

In FIG. 2, reference numeral 21 denotes an image input device, such as a scanner or a digital camera, for generating image data to be input according to the present invention. 27 is an instruction input device for inputting each instruction of the system user, and mainly uses a keyboard and a mouse. Reference numeral 22 denotes a network cable line for connecting each device to form a network, and is typically an Ethernet cable currently used for the global Internet or a local intranet.

Reference numeral 23 denotes a video driver which performs data control for displaying image data from each input device 21 and image data via the network cable line 22. 2
Reference numeral 4 denotes an image output device for processing and displaying the received image data, and has a built-in video driver 23. Specific examples include a personal computer and a large screen monitor.

Reference numeral 25 denotes a printer driver which performs image data conversion for outputting image data from each input device 21 and image data via the network cable line 22 to paper. Reference numeral 26 denotes a printer that outputs to paper based on the image data converted by the printer driver 25.

The image processing apparatus for performing the resolution conversion processing according to the present invention is built in, for example, the video driver 23 and the printer driver 25, and operates when each image data is converted.

FIG. 3 shows another example of the system configuration in which the image processing apparatus according to the present invention is employed. This configuration realizes pattern matching between a template image and an input image, and includes, for example, an imprint collation system.

In FIG. 3, reference numeral 31 denotes an image input device for generating image data to be input according to the present invention, and a scanner is mainly used. Reference numeral 32 denotes an instruction input device for inputting each instruction of the system user, and mainly uses a keyboard.

Reference numeral 33 denotes a template file for storing, as template data, characteristics of reference imprints serving as basic image data for pattern matching. Reference numeral 34 denotes a template file 33 based on an instruction from the instruction input device 32.
Is a template reading unit for selecting desired template data from.

An image adjustment unit 35 performs various adjustments between target images necessary for matching the input image data obtained by the image input device 31 with the template data obtained from the template file 33. Image adjustment mainly includes image enlargement / reduction and image rotation.

Reference numeral 36 denotes a normalization unit for extracting features suitable for image comparison from the target image data obtained by the image adjustment unit 35. Reference numeral 37 denotes a matching unit that matches target image data obtained by the normalizing unit 36 with corresponding template data obtained through the template reading unit 34. Numeral 38 denotes a result output unit for indicating the collation result obtained by the collation unit 37 to the user. 39 is a result output device for displaying the output from the result output unit 38.

The image processing apparatus for performing the resolution conversion processing and the like according to the present invention is built in, for example, the image adjustment unit 35 and operates particularly when converting the resolution of the input image data at the time of enlargement.

The image processing according to the present invention may be realized by hardware or software.

Next, the operation of each component in the image processing apparatus of this embodiment shown in FIG. 1 will be described in detail.

FIG. 4 is a diagram showing an example of a method of calculating a filter coefficient for enhancing an edge in the enhancement coefficient determination unit 3 shown in FIG. Here, x is a variable determined by the number of interpolation pixels input from the input terminal 2. In the present embodiment, the variable x is given so as to be proportional to the number of interpolation pixels.

When it is desired to emphasize the edges in the horizontal and vertical directions, the center is set to x + 1, the value on the diagonal line is set to -x / 4, and the other values are set to 0, as shown in FIG. Calculate the coefficient value. When it is desired to emphasize the diagonal edge, the center is set to x + 1 as shown in FIG.
Each coefficient value is calculated by setting the upper and lower values to -x / 4 and setting the other values to 0.

Next, the operation of the edge enhancement processing in the edge enhancement image generator 4 will be described. In the target pixel P _{i, j} of the original low-resolution image, a pixel subjected to edge enhancement processing corresponding to the same pixel position using the filter shown in FIG.
_{Assuming that i, j} , the edge emphasis pixel E _{i, j} is obtained by the following equation using four peripheral pixels.

E _{i, j} = (x + 1) × P _{i, j} −x × (P _{i−1, j−1}
+ P _{i-1, j + 1} + P _{i + 1, j-1} + P _{i + 1, j + 1} ) / 4 Similarly, when the filter shown in FIG. 4B is used, Ei, j = ( x + 1) × P _{i, j} −xx × (P _{i, j-1} + P _{i-1, j} + P
_{i, j + 1} + P _{i + 1, j} ) / 4.

When this filter is used, the absolute value of the edge-enhanced pixel becomes larger as the difference between the density values of the target pixels used for calculation becomes larger, and as the difference becomes smaller, the edge-enhanced pixel becomes smaller. The absolute value has the characteristic of becoming smaller. A positive value is obtained for a pixel having a high density at the edge, and a negative value is obtained for a pixel having a low density at the edge.

Next, the operation of the interpolation pixel generator 5 will be described. For the target pixel position (i, j) of the enhanced image output from the edge enhanced image generation unit 4, an interpolation pixel is generated using a bilinear method or a cubic convolution method, which is a conventional interpolation pixel generation method. The pixel generation method according to each of these methods has been described in the related art, and will not be described here.

FIG. 5 is a diagram showing how pixels of low resolution information after edge enhancement are generated as interpolation pixels by the bilinear method. In the drawing, for simplicity of description, only the one-dimensional direction in the main scanning direction or the sub-scanning direction is shown as an object. In the figure, the horizontal direction indicates the pixel position, and the vertical direction indicates the brightness or density at each pixel position.

That is, the position and density of one pixel in one block are shown. Here, two adjacent pixels, which are edges, are targeted. Pixels indicated by thick frames in the figure indicate the original low resolution information. The halftone portion is a pixel density emphasized by the edge-enhanced image generation unit 4. According to the characteristics of the filter described above, the higher the edge density is, the higher the edge density is, and the lower edge portion is lower by the edge enhancement process. The density value changes.

The thick solid line in the figure indicates that the interpolation pixel is generated by applying the bilinear method to the edge emphasized image.
A broken line in the drawing indicates a density value when an interpolated pixel is generated by applying the bilinear method or the nearest neighbor method to the original low-resolution image without edge enhancement.

Next, the operation of the comparing section 6 will be described.

In the comparing section 6, the maximum density value used as a threshold value in the edge processing of the present embodiment is selected from the peripheral pixels at the pixel position of interest (i, j) specified by the input value of the information input terminal 9. And the minimum density value. That is, assuming that the value of the peripheral pixel range obtained by the information input terminal 9 is n, the comparing unit 6 determines that the pixel position of interest (i, j) and the neighboring pixel positions (i-n + 1, j) (i, j−n + 1) (i−
n + 2, j) (i, j-n + 2) ... (i + n, j +
A maximum density value and a minimum density value are detected from the pixel n).

FIG. 16 shows peripheral pixel ranges when n = 1, 2, and 3. The value of n that determines the peripheral pixel range is
When the input image is in focus, the input image is low, and when the image is out of focus, the input image is increased. here,
An example where n = 1 is described.

In this case, the pixel position (i, j), (i + 1, j), (i, j + 1), (i +
A maximum density value and a minimum density value are detected from the four pixels (1, j + 1). _Assuming that the maximum density value is P _max and the minimum density value is P _min , that is, P _max = max (P _{i, j} , P _{i + 1, j} , P _{i, j + 1} , P
_{i + 1, j + 1} ) P _min = min (P _{i, j} , P _{i + 1, j} , P _{i, j + 1} , P
_{i + 1, j + 1} ).

The target pixel position (i, j) in the original low-resolution information is processed by the edge emphasized image generation unit 4 and the interpolation pixel generation unit 5 to generate M × M interpolation pixels including the original pixel. Suppose you did. In the present embodiment, all of the M × M interpolated pixels are compared with P _max and P _min, and among the interpolated pixels, P
_The density value of a pixel position having a density value larger than _max is P
the _max, and outputs the weighting factor to replace the density value of the pixel positions having a P _min smaller density value P _min.

The weighting / synthesizing section 7 weights the low-resolution pixels and the output pixels of the interpolation pixel generating section 5 based on the weighting coefficients output from the comparing section 6. Specifically, the interpolation pixel P ′ _{i + Di, j + Dj} at the output terminal 8 of the interpolation pixel position value (i + Δi, j + Δj) is P ′ _{i + Di, j + Dj} = W _max × P _max + W _min × P _min + (1-W
_max ) (1−W _min ) × P _{i + Di, j + Dj} .

Here, Di and Dj mean Δi and Δj,
W _max and W _min are respectively W _max = 1, W _min = 0 when _{Pi + Di, j + Dj} > P _max , and W _max = 0 when _{Pi + Di, j + Dj} <P _{min Wmin} = 1, _Pmin < _{Pi + Di, j + Dj} < _Pmax , _Wmax = 0, W
It is a weight coefficient given so that _min = 0.

FIG. 6 shows the value of each pixel when these processes are performed. In this drawing, as in FIG. 5, only the one-dimensional direction in the main scanning direction or the sub-scanning direction is shown. Here, it is assumed that the density values of two adjacent pixels, which are the original low-resolution information indicated by the thick frames, are P _max and P _min , respectively.

According to the present embodiment, among the pixels interpolated by the bilinear method based on the edge-emphasized pixels, pixels exceeding the original pixel density P _max are replaced with P _max , and the original pixel density P _min 5 are replaced with P _min , and FIG.
From the interpolation result, an interpolation pixel along the thick solid line in FIG. 6 is newly obtained.

FIG. 7 shows an example of interpolation pixel generation using the edge processing method of the present embodiment.

FIG. 7A shows low resolution information as original pixels. Pixels obtained by performing edge enhancement on this information are shown in (b)
Then, an interpolation pixel is generated by using the bilinear method or the cubic convolution method, and (c) is obtained.

For each pixel in (c), the corresponding pixel density value in (a) is compared, and the density value of the pixel position having a density value larger than the original pixel and the pixel having a density value smaller than the original pixel are compared. The density value at the position is replaced with the maximum value and the minimum value of the density value of the original pixel, and (d) is obtained.

With the above processing method, it is possible to convert low-resolution information into high-resolution information reflecting characteristics of the density value and density value distribution.

FIG. 8 is a diagram showing a series of these operations in terms of specific pixel density values. In the figure, the number of 2 × 2 pixels is 4 ×
An example is shown in which the number of pixels is converted to four. (A) in the figure
The pixel position indicated by the bold frame is a pixel to be interpolated as an original pixel. It is assumed that the filter shown in FIG. 4A is used in the emphasis coefficient determination unit 3, and in this case, x = 2 is determined.

The edge-enhanced pixel value obtained in the edge-enhanced image generation unit 4 using this filter is represented by (b) in FIG.
Shown in (C) in the figure shows a pixel value when an interpolation pixel is generated by the bilinear method in the interpolation pixel generation unit 5 for each pixel value. _Pmax and _Pmin are obtained from the pixel values shown in FIG. In this case, P _max = 2
00, P _min = 1.

When the weighting / synthesizing unit 7 compares the values of P _max and P _min with respect to each pixel in FIG. 3C and compares them with each other, and executes the replacement, the pixel values in FIG. Desired.

The high-resolution information obtained by the weighting / synthesizing section 7 is output from an output terminal 8 to an image output device such as a printer or a monitor.

According to the present embodiment, an image processing apparatus and an edge processing method capable of suppressing blurring of an edge area after high-resolution conversion without changing the color scheme of an original image. Can be provided.

According to the image display / printing system as shown in FIG. 2 which employs the image processing apparatus of the present embodiment, even when the original image is enlarged and resized, the color arrangement characteristic of the original image is obtained. Is not changed, and the occurrence of blur in the edge region can also be suppressed.

Further, according to the seal imprint collation system as shown in FIG. 3 employing the image processing apparatus of the present embodiment, the input image data to be collated is reduced, and this is enlarged to generate template data and When comparing and matching,
Since the occurrence of blurring in the edge region and a change in color arrangement, which are likely to occur when the input image is enlarged, are suppressed, the matching accuracy can be improved.

In this embodiment, the variable x in FIG. 4 is given in proportion to the number of pixels to be interpolated. Alternatively, the density value distribution around the pixel to be interpolated obtained from the image input terminal 1 may be used. May be determined according to the characteristics of

For example, a configuration may be adopted in which the frequency distribution around the pixel to be interpolated is examined, and the value of x is reduced when a large number of high-frequency distributions are found, and is increased when not.

In the comparison section of the present embodiment, the maximum density value and the minimum density value of the peripheral pixel group existing around the target pixel are obtained as threshold values, but the threshold value can be used in the present invention. The values are not limited to these. For example, a new threshold value is obtained from the maximum density value and the minimum density value, or the color change (density value distribution) in the peripheral pixel group is added to the maximum and minimum density values, A configuration may be adopted in which directions and the like are extracted and threshold values are set so that these are reflected.

In this embodiment, an example in which the present invention is applied to the resolution conversion processing has been described. However, the present invention is applicable to any image processing other than the resolution conversion processing that involves edge processing. For example, as in the present embodiment,
The occurrence of blur in the edge region can be suppressed.

For example, edge enhancement processing is performed on an input image or an image obtained by performing image processing on an input image by a predetermined method, or predetermined image processing is performed on the edge-enhanced image after performing the edge enhancement processing. The at least edge-enhanced image is compared with the input image and the emphasized image so as not to change a color scheme or shading characteristic determined by a density value distribution in an edge region of the input image according to the comparison result. The configuration may be such that the density value of the pixel in the edge region is reset.

(Embodiment 2) FIG. 17 shows a schematic block diagram of an image processing apparatus according to the present embodiment. The image processing apparatus according to the present embodiment performs image processing on an image whose edge region is blurred, such as an out-of-focus image, so that the blurred region can be removed and the edge can be sharpened. 1701 to 1709 described below.
Is provided.

In FIG. 17, reference numeral 1701 denotes an image information input terminal for inputting original image information to be subjected to image processing. Generally, an image data input terminal is a scanner, a digital camera, image data displayed on a screen, or the like. Data. 1
Reference numeral 702 denotes an edge enhancement coefficient input terminal for inputting information indicating the degree of edge enhancement.

Reference numeral 1704 denotes an edge-enhanced image generation unit for edge-enhancing the input image obtained from the input terminal 1701 using the determined enhancement coefficient. A comparison unit 1706 compares the target pixel of the input image input from the image input terminal 1701 with the output value of the edge enhanced image generation unit 1704, and outputs a value indicating the degree of weighting of each image value according to the result. It is.

Reference numeral 1709 denotes an information input terminal for giving information for arbitrarily specifying a peripheral pixel range for detecting the maximum density value and the minimum density value to the comparison unit 1706.

Reference numeral 1707 denotes a pixel of interest of the input image input from the image input terminal 1701 and an edge-enhanced image generation unit 1704 according to the weighting output from the comparison unit 1706.
Is a weighting synthesis unit that weights and synthesizes the output values of.

Reference numeral 1708 denotes an image output terminal for outputting information of an output image subjected to image processing by the image processing apparatus to the outside. The output of the weighting / synthesizing unit 1707 is output.

FIG. 18 shows an example of a system configuration in which the image processing apparatus of this embodiment is adopted. This configuration is a device that stores a two-dimensional image obtained by a non-contact input device as data and outputs the data to a display or paper.

In FIG. 18, reference numeral 1801 denotes an image input device for generating image data to be input according to the present invention.
Digital cameras and non-contact scanners are mainly used. Reference numeral 1802 denotes an instruction input device for inputting each instruction of the user of the present system, and mainly uses a keyboard.

Reference numeral 1804 denotes an image adjustment unit for adjusting the image of the input image data obtained by the image input device 1801. Image adjustment mainly includes surface correction and blur removal of an image for conversion into a good two-dimensional image. Reference numeral 1805 denotes a data storage unit that stores the image data adjusted by the image adjustment unit 1804.

Reference numeral 1806 denotes an image output device for displaying data in the data storage unit 1805 on a screen. I8
Reference numeral 07 denotes a printer for outputting data in the data storage unit 1805 to paper.

The image processing apparatus according to the present embodiment is built in, for example, the image adjustment unit 1804 and functions as a blur removal processing for a blurred image.

The operation of each unit in FIG. 17 is different from that of FIG. 1 in that the interpolation pixel generation unit 5 and the enhancement coefficient determination unit 3 are eliminated and an edge enhancement coefficient input terminal 1702 for directly inputting information of the degree of edge enhancement is provided. Performs the same operation as the image processing apparatus in FIG.

That is, the input original image is edge-enhanced by the edge-enhanced image generation unit 1704. Comparison section 1
At 706, the edge-enhanced image is compared with the original image, and according to the comparison result, the density value of each pixel of the edge-enhanced image does not exceed the maximum and minimum density values of the original image, respectively. Weighting is performed on each pixel of the edge-emphasized image. In accordance with the weighting, the weighting / synthesizing unit 1707 synthesizes the original image and the edge-emphasized image, and obtains an output image in which the blurred portion is reduced. Since the detailed operation of each unit has been described in the first embodiment, it will not be described here.

However, in the present embodiment, the edge enhancement coefficient x given by the edge enhancement coefficient input mizuko l702 is
It is arbitrarily given by the user, and a larger integer value is given as the user wants to strengthen the edge emphasis.

The value n indicating the peripheral pixel range of the target pixel for detecting the maximum density value and the minimum density value, which is input from the information input terminal l709 and used by the comparison unit 17O6, is in focus on the input image. It is preferable that the image is low and the image is out of focus, so that the image is enlarged. Specifically, it is assumed that the user gives the maximum density value and the minimum density value from outside the blurred area of the edge.

According to the image processing apparatus of the present embodiment or a system employing the same, by performing the same processing as in the first embodiment, for example, as shown in FIG.
It is possible to obtain an output image in which the blur has been eliminated from the input image in which the blur has occurred.

[0100]

As described above, according to the present invention, it is possible to provide a resolution conversion processing device in which edges are emphasized and the coloration of input low-resolution information is not changed.

Further, according to the present invention, even when an out-of-focus image or an image with a blurred edge is given as an input image, the edge is enhanced by adjusting the edge enhancement coefficient and the peripheral pixel range. In addition, it is possible to obtain an image which does not change the color scheme of the input original image and which is free from blur.

[Brief description of the drawings]

FIG. 1 is a configuration block diagram of an image processing apparatus according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram showing an example of the overall configuration of an image display and printer system in which the image processing apparatus of the present invention is employed.

FIG. 3 is an explanatory diagram showing an example of the overall configuration of a seal imprint collation system that employs the image processing apparatus of the present invention.

FIG. 4A is an explanatory diagram showing an example of an edge enhancement filter according to the present invention. FIG. 4B is an explanatory diagram showing another example of the edge enhancement filter of the present invention.

FIG. 5 is an explanatory diagram one-dimensionally illustrating an example in which an interpolation pixel is generated by a bilinear method after edge enhancement processing.

FIG. 6 is an explanatory diagram one-dimensionally showing an example of generating an interpolation pixel according to the present invention.

FIG. 7A is an explanatory diagram two-dimensionally showing a pixel state when an interpolation pixel is generated according to the present invention. FIG. 7 (b): The pixel state at the time of generating the interpolation pixel according to the present invention is 2
FIG. FIG. 7C: The pixel state at the time of generating the interpolation pixel according to the present invention is 2
FIG. FIG. 7D: The pixel state at the time of generating the interpolation pixel according to the present invention is 2
FIG.

FIG. 8A is an explanatory diagram showing specific pixel values of a pixel state when an interpolation pixel is generated according to the present invention. FIG. 8B is an explanatory diagram showing a pixel state at the time of generating an interpolation pixel by specific numerical values according to the present invention. FIG. 8C is an explanatory diagram showing the pixel states at the time of generating the interpolation pixels by specific numerical values according to the present invention. FIG. 8 (d) is an explanatory diagram showing a pixel state at the time of generating an interpolation pixel by specific numerical values according to the present invention.

FIG. 9 is an explanatory diagram showing the principle of the nearest neighbor method.

FIG. 10 is an explanatory diagram illustrating the principle of the bilinear method.

FIG. 11 is an explanatory diagram showing the principle of the cubic convolution method.

FIG. 12 is an explanatory diagram showing an example of pixel generation by the nearest neighbor method.

FIG. 13 is an explanatory diagram showing an example of pixel generation by a bilinear method.

FIG. 14 is an explanatory diagram illustrating a configuration example of an edge enhancement filter.

FIG. 15A is an explanatory diagram showing an example of pixel generation by the bilinear method after edge enhancement. FIG. 15B is an explanatory diagram showing an example of pixel generation by the bilinear method after edge enhancement. FIG. 15C is an explanatory diagram showing an example of pixel generation by the bilinear method after edge enhancement.

FIG. 16 is an explanatory diagram illustrating an example of a peripheral pixel range for detecting a maximum density value and a minimum density value.

FIG. 17 is a configuration block diagram of an image processing apparatus according to another embodiment of the present invention.

FIG. 18 is an explanatory diagram illustrating an example of the overall configuration of an image processing system that employs the image processing apparatus of the present invention.

FIG. 19A is an explanatory diagram illustrating an example of pixel generation by edge processing according to the second embodiment. FIG. 19B is an explanatory diagram illustrating an example of pixel generation by edge processing according to the second embodiment.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Image input terminal, 2 ... Information input terminal, 3 ... Enhancement coefficient determination part, 4 ... Edge enhancement image generation part, 5 ... Interpolation pixel generation part, 6 ... Threshold value determination part, 7 ... Pixel determination part, 8 ... Image output terminal, 9: Information input terminal, 21: Image input device, 22
... Network cable line, 23 ... Video driver, 2
4 image output device, 25 printer driver, 26 printer device, 31 image input device, 32 instruction input device, 33 template file, 34 template read unit, 35 image adjustment unit, 36 normalization unit , 37
... A collating unit, 38 a result output unit, 39 a result output device.

Continued on the front page (72) Inventor Toshiaki Nakamura 7-1-1, Omika-cho, Hitachi City, Ibaraki Prefecture Inside Hitachi Research Laboratory, Hitachi, Ltd. (72) Keisuke Nakajima 7-1-1, Omika-cho, Hitachi City, Ibaraki Prefecture F term in Hitachi Research Laboratory, Hitachi, Ltd. (Reference) 5B057 BA02 BA25 CA01 CA02 CA08 CA12 CA16 CB01 CB02 CB08 CB12 CB16 CC01 CD06 CE03 CE08 DA20 DB02 DB05 DB06 DB09 DC33 5C076 AA12 AA21 AA22 BB04 5C077 PP03 PP55 02

Claims (13)

[Claims] An image processing apparatus for performing resolution conversion by interpolating pixel values of a multi-valued image as an input image and interpolating pixel values of the input image, comprising: an image input unit for inputting the input image information; Edge-enhanced image generation means for emphasizing; interpolated pixel generation means for generating an interpolated pixel by interpolating a pixel value at a specified position on the edge-emphasized input image; From a pixel value of a peripheral pixel, a threshold value is obtained by a predetermined method, and a comparison unit that compares the threshold value with the pixel value of the interpolation pixel generated by the interpolation pixel generation unit, In response, weighting is performed on the interpolated pixel values, and a weighting / synthesizing unit that synthesizes higher-resolution output image information is output to the outside. The image processing apparatus characterized by having an image output means for. 2. The image processing apparatus according to claim 1, wherein: information input means for inputting information on the pixel value interpolation processing; and information input from the image input means and the information input means. Enhancement coefficient determining means for determining a coefficient of a filter used for edge enhancement processing, wherein the edge-enhanced image generation means performs edge enhancement on the input image information using the determined filter coefficient. Image processing device. 3. The image processing apparatus according to claim 2, wherein the emphasis coefficient determining unit changes the filter coefficient according to a pixel value of a peripheral pixel at a specified position on the input image. Characteristic image processing device. 4. The image processing apparatus according to claim 2, wherein the information input means receives an input of the number of interpolation pixels to be interpolated, and the emphasis coefficient determination means determines the number of interpolation pixels to be inputted. An image processing apparatus, wherein the filter coefficient is changed in proportion. 5. The image processing apparatus according to claim 1, wherein said comparing means obtains a maximum and a minimum density value as a threshold value in a group of peripheral pixels at a designated position on said input image. Comparing the threshold value with the generated interpolation pixel so as not to exceed the maximum and minimum density values,
A weighting coefficient for resetting each pixel value in the generated group of interpolated pixels is determined.The weighting / combining unit combines the output image information according to the obtained weighting coefficient. Characteristic image processing device. 6. The image processing apparatus according to claim 1, wherein said comparing means can arbitrarily set a range of peripheral pixels used for determining said threshold value. . 7. A printing system having at least a printer driver for converting input image data into data and a printer for outputting data using the converted image data, wherein the printer driver comprises: A printing system comprising at least the image processing device according to any one of claims 6 to 10. 8. An image display system comprising: a video driver for performing data control for displaying input image data; and an image output device for processing and displaying the image data. Item 7. An image display system comprising at least the image processing device according to any one of Items 1 to 6. 9. An image input device for generating image data to be input, an instruction input device for inputting each instruction of a user of the system, and a feature of a reference imprint serving as basic image data for pattern matching as a template. A template file that is stored as an image, a template reading unit that selects a desired template image from the template file based on an instruction from the instruction input device, and an input image obtained by the image input device and a template file that is obtained from the template file. An image adjustment unit that performs adjustment between target images necessary for matching with the template image, a normalization unit that extracts target image data obtained from the image adjustment unit and extracts features suitable for image matching, The target image data obtained by the normalization unit and the corresponding template obtained through the template reading unit. A seal imprint collation system having a collating unit for collating a plate image, a result output unit for presenting a collation result obtained by the collating unit to a user, and a result output device for displaying an output from the result output unit In this embodiment, the image adjustment unit includes at least the image processing device according to any one of claims 1 to 6. 10. An image processing apparatus for processing a multi-valued image as an input image and performing edge processing on the input image, wherein: an image input means for inputting the input image information; and an image in which the input image information is edge-emphasized. An edge-enhanced image generating means, and a threshold value is obtained from a pixel value of a peripheral pixel at a designated position on the input image by a predetermined method, and the threshold value is generated by the edge-enhanced image generating means. Comparing means for comparing pixel values of the image; weighting combining means for combining output image information by performing weighting on the pixel values of the edge-emphasized image according to the comparison result; and An image output unit that outputs the output image information synthesized by the synthesis unit to the outside. 11. An image input device for generating image data to be input, an instruction input device for inputting each instruction of a user of the present system, and the image input device based on the instruction from the instruction input device. An image adjustment unit that performs image processing on the input image obtained by the above, a data accumulation unit that accumulates image-adjusted image data output from the image adjustment unit, and outputs the image data accumulated in the data accumulation unit An image processing system having an image output device, wherein the image adjustment unit includes at least the image processing device according to claim 10. 12. An image processing apparatus for generating information indicating an output image by performing at least edge processing on an input multi-valued image, wherein at least edge enhancement processing is performed on the input image using the input image. First processing means for performing image processing; and determining a threshold value by a predetermined method using a pixel value in a pixel group including at least a pixel to be subjected to the edge enhancement processing in the input image. Comparing means for comparing a threshold value with a pixel value of the pixel group subjected to the edge enhancement processing in the image output from the first image processing means; and, in accordance with the comparison result, the first image processing means Second processing means for resetting the pixel values of the pixels subjected to the edge enhancement processing in the image output from the image processing apparatus, and synthesizing the output image according to the reset pixel values. The image processing apparatus characterized by. 13. An edge processing method for a multi-valued image, wherein image processing including at least edge enhancement processing is performed on the multi-valued image, and a periphery including at least a pixel to be subjected to the edge enhancement processing in the input image. An edge, wherein a feature amount is obtained with respect to a pixel value distribution of a pixel group, and a pixel value of a pixel existing in the edge-emphasized region in the image-processed image is reset according to the feature amount. Processing method.