JP2010191931A - Image processing apparatus, image forming apparatus, computer program, recording medium and image processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing apparatus, an image forming apparatus, a computer program, a recording medium, and an image processing method, to achieve a smoothing process with reduced jaggy without causing a blur at a contour portion due to an interpolation process. <P>SOLUTION: A same attribute layer extraction part 2121 extracts the same attribute layer information indicating a pixel region of the same feature amount as a pixel of interest in the periphery of the interpolation target coordinates. A layer information interpolation part 2122 determines an interpolated value of the same attribute layer information in each of the interpolation target coordinates. A layer information filter part 2123 performs low pass filter processing to the interpolation layer information to output the degree of impact. A thin layer extraction part 2124 determines whether the pixel of interest be a thin line or not. A copy origin pixel selection part 2125 selects a copy origin pixel correlated to each of the interpolation target coordinates. A pixel feature amount copying part 2126 copies a pixel value (feature amount) of a selected copy original pixel correlated to each of the interpolation target coordinates. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image processing apparatus that performs high-resolution processing of a color multi-value image and performs smoothing processing on the contour of a drawing element, an image forming apparatus including the image processing apparatus, and a computer program for realizing the image processing apparatus The present invention relates to a recording medium on which the computer program is recorded and an image processing method.

  There are various interpolation methods such as nearest neighbor interpolation (nearest neighbor interpolation), linear interpolation (bilinear interpolation), and nonlinear interpolation (bicubic, etc.) in order to increase the resolution of an image. Nearest neighbor interpolation is the simplest method, and is an interpolation method in which the pixel value (feature value) of the pixel closest to the interpolation target coordinate is used as it is as the pixel value of the interpolation pixel in the interpolation target coordinate. FIG. 60 is an explanatory diagram showing a part of an original image, and FIG. 61 is an explanatory diagram showing a part of an image after processing by nearest neighbor interpolation as a conventional example. In the nearest neighbor interpolation, as shown in FIG. 61, only the resolution is increased while maintaining the same shape as the original image as shown in FIG. Without contrast, it is possible to maintain the contrast of the outline, which is important in artificial images such as characters and vector graphics, but diagonal lines and curved jaggy (stepped jagged lines appearing when diagonal lines are expressed) There is a disadvantage that stands out.

  On the other hand, in linear interpolation or nonlinear interpolation, it is possible to reduce jaggy of diagonal lines and curves compared to nearest neighbor interpolation. FIG. 62 is an explanatory diagram showing a part of an image after processing by linear interpolation as a conventional example. Linear interpolation is suitable for increasing the resolution of natural images such as photographs because the density of interpolated pixels changes stepwise, but for artificial images such as characters and vector graphics, drawing is possible. There is a drawback that the outline of the element is blurred.

  Several interpolation methods have been proposed as methods for solving such problems. For example, there is an image processing method that can improve the image quality of an interpolated image by determining an image attribute in a specified region around the coordinates to be interpolated and selecting an optimal interpolation method or interpolation parameter according to the determined image attribute. It is disclosed (see Patent Document 1).

  In addition, for jaggy images such as the image after nearest neighbor interpolation, the feature values of the target pixel and the peripheral pixels are compared with each other, and the number of pixels having the same feature value is obtained for each of the target pixel and the background pixel. An image processing method that performs smoothing processing by replacing a target pixel with a feature amount of a background pixel by combining the number of pixels is disclosed (see Patent Document 2).

In addition, smoothing processing is performed on contour information obtained by pattern matching or edge extraction based on binary foreground / background information obtained based on region information indicating whether a pixel belongs to a text region or a character region. An image processing method is disclosed in which smoothing is performed on multi-valued color images without causing blurring of edges by copying representative values of foreground / background regions based on the results (Patent Document 3). reference).

JP 2000-339449 A JP 2005-192184 A JP 2000-125134 A

  However, the image processing method of Patent Document 1 shows a method of selecting a more suitable interpolation method with reference to surrounding pixel information, but the interpolation method itself is a combination of conventional methods, The point of blurring in the part has not been improved.

  Further, in the image processing method of Patent Document 2, since only the peripheral pixels of the target pixel are processed with respect to the high resolution image, there is only a local smoothing effect, and an oblique line with a gentle angle is formed. On the other hand, there is little smoothing effect. FIG. 63 is an explanatory diagram showing a part of an image after interpolation by comparison of neighboring pixels as a conventional example. The example of FIG. 63 shows an example when the image processing method of Patent Document 2 is performed on the original image shown in FIG. In this case, there is a drawback in that the higher the resolution ratio, the less the smoothing effect against jaggies.

  Further, in the image processing method of Patent Document 3, smoothness can be obtained with respect to the contour shape of the drawing element, and the representative value of the drawing element can be copied as it is. Since tag information as contour information is required, there is a limitation that it cannot be applied to an image without TAG information such as a TIFF image. In addition, since the boundary between drawing elements having the same tag information such as overlapping of characters of different colors cannot be determined, it cannot be applied to such a boundary portion.

  The present invention has been made in view of such circumstances, and performs high-resolution color multi-valued images, and performs smoothing processing that reduces jaggies without causing blurring due to interpolation processing in the contour portion. An object of the present invention is to provide an image processing apparatus, an image forming apparatus including the image processing apparatus, a computer program for realizing the image processing apparatus, a recording medium storing the computer program, and an image processing method. .

  The image processing apparatus according to the present invention is an image processing apparatus that generates an output image by performing an interpolation process on an input image. The degree of influence of each of a plurality of pixels of interest around a plurality of interpolation target coordinates at the interpolation target coordinates In accordance with the influence degree calculated by the influence degree calculating means, a set of attention pixel groups having the same influence degree is selected from the attention pixels in association with each interpolation target coordinate according to the influence degree calculated by the influence degree calculating means. And a interpolation pixel value calculation unit that calculates a pixel value of the interpolation pixel of the interpolation target coordinate based on the pixel value of the target pixel group selected by the selection unit.

  The image processing apparatus according to the present invention includes, for each target pixel, the same attribute determination unit that determines the presence / absence of the same attribute between the target pixel and each pixel in the pixel block including the plurality of target pixels, and the same attribute determination Extracting means for extracting the same attribute information assigned in correspondence with the respective pixels, and same attribute information interpolating means for interpolating the attribute information extracted by the extracting means at a predetermined interpolation magnification. And the influence degree calculating means is configured to calculate the influence degree due to neighboring pixels existing around the coordinates to be interpolated by using the same attribute information interpolated by the same attribute information interpolation means for each target pixel. It is characterized by being.

  In the image processing apparatus according to the present invention, the same attribute determination unit determines whether or not the same attribute exists for each target pixel depending on whether or not the pixel values of the target pixel and each pixel in the pixel block match. It is characterized by being comprised so that it may determine.

  In the image processing apparatus according to the present invention, the attribute determination unit includes, for each target pixel, a pixel value of each pixel in the pixel block included in a pixel value range provided for the pixel value of the target pixel. It is characterized in that it is configured to determine the presence or absence of the same attribute according to whether or not it is.

  In the image processing apparatus according to the present invention, the attribute determination unit selects, for each target pixel, the pixel value range based on the image attribute value of the target pixel, and whether the pixel value of the target pixel is included in the pixel value range. Whether or not the image attribute value is the same is determined based on whether or not the image attribute value is the same, and the interpolated pixel value calculating unit is configured to determine whether or not the pixel group of interest selected by the selecting unit is selected. The pixel value of the interpolation pixel is calculated from the pixel value according to a procedure selected according to the image attribute value of the target pixel group.

  In the image processing apparatus according to the present invention, the same attribute determination unit is configured to detect, for each target pixel, each pixel in the pixel block adjacent to the target pixel via the target pixel or a pixel that has already been determined to have the same attribute. It is configured to determine the presence or absence of the same attribute according to whether or not to do so.

  The image processing apparatus according to the present invention is characterized in that the selection unit is configured to select a pixel of interest having the greatest influence calculated by the influence calculation unit for each interpolation target coordinate. .

  The image processing apparatus according to the present invention includes first comparison means for comparing the influence degree calculated by the influence degree calculation means with a first threshold value, and the selection means has the influence degree at the interpolation target coordinates as the first degree. When the value is smaller than one threshold, the pixel of interest closest to the interpolation target coordinate is selected.

  The image processing apparatus according to the present invention includes a thin line determination unit that determines whether or not a target pixel belongs to a diagonal line or a thin line that is an isolated point, and a target pixel that is determined to belong to a thin line by the thin line determination unit Second comparison means for comparing the degree of influence of the interpolation target coordinates with a second threshold value, and the selection means, when the degree of influence of the interpolation target coordinates is greater than the second threshold value, The pixel of interest determined to belong to the thin line is selected in association with the coordinate, and when the degree of influence at the interpolation target coordinate is smaller than the second threshold value, the pixel of interest is associated with the interpolation target coordinate. The pixel of interest having the greatest influence is configured to be selected.

  The image processing apparatus according to the present invention includes a thin line determination unit that determines whether or not a target pixel belongs to a diagonal line or a thin line that is an isolated point, and a target pixel that is determined to belong to a thin line by the thin line determination unit Second comparison means for comparing the degree of influence of the interpolation target coordinates with a second threshold value, and the selection means, when the degree of influence of the interpolation target coordinates is greater than the second threshold value, The pixel of interest determined to belong to the thin line is selected in association with the coordinate, and the first comparison unit, when the influence degree at the interpolation target coordinate is smaller than the second threshold, The influence degree calculated by the influence degree calculating means is configured to be compared with the first threshold value. When the influence degree at the interpolation target coordinate is smaller than the first threshold value, the attention closest to the interpolation target coordinate is obtained. The pixel is selected, If impact in serial interpolation target coordinates is equal to or greater than the first threshold value, characterized in that the degree of influence calculated by the influence degree calculation means is arranged to select the highest pixel of interest.

  In the image processing apparatus according to the present invention, when the thin line determination unit determines that the same attribute determination unit does not have the same attribute of the pixel of interest and a pixel adjacent to the pixel of interest in the vertical or horizontal direction, It is constituted so that it may be judged that it belongs to.

  The image processing apparatus according to the present invention includes a changing unit that changes the second threshold according to the number of pixels having the same attribute as the target pixel determined to belong to the thin line by the thin line determining unit.

  An image forming apparatus according to the present invention includes an image forming unit that forms an output image, and an image processing apparatus according to any one of the above-described inventions.

  A computer program according to the present invention is a computer program for causing a computer to function as means for generating an output image by performing an interpolation process on an input image. The computer includes a plurality of pixels of interest around a plurality of interpolation target coordinates. A degree of influence calculating means for calculating the degree of influence at each of the interpolation target coordinates, and a set of attentions having the same degree of influence from the target pixel in association with each interpolation target coordinate according to the calculated degree of influence A selection unit that selects a pixel group; and an interpolation pixel value calculation unit that calculates a pixel value of an interpolation pixel at an interpolation target coordinate based on a pixel value of a target pixel group selected by the selection unit. .

  A recording medium readable by a computer according to the present invention records the computer program according to the above-described invention.

  An image processing method according to the present invention is an image processing method by an image processing apparatus that generates an output image by performing an interpolation process on an input image, and the interpolation target of each of a plurality of pixels of interest around a plurality of interpolation target coordinates The degree of influence at the coordinates is calculated, and in accordance with the calculated degree of influence, a set of target pixel groups having the same degree of influence is selected from the target pixels in association with each interpolation target coordinate and selected. The pixel value of the interpolation pixel of the interpolation target coordinate is calculated based on the pixel value of the target pixel group.

  In the present invention, the degree of influence at the interpolation target coordinates of each of the plurality of target pixels around the plurality of interpolation target coordinates is calculated. For example, it is assumed that the resolution is increased by paying attention to a 2 × 2 pixel grid (target pixel) in the input image and obtaining an interpolation pixel in the 2 × 2 pixel grid. The number of interpolation pixels is determined by the interpolation magnification that is the resolution ratio between the input image and the output image. In this case, the degree of influence at each interpolation target coordinate is calculated for each of the four target pixels. The degree of influence can be obtained, for example, according to how much pixels having the same pixel value are distributed to reference pixels around the target pixel. In accordance with the calculated influence level, a set of target pixel groups having the same influence level is selected from the target pixels in association with each interpolation target coordinate. For example, assuming that the (n, m) coordinate in the 2 × 2 pixel grid is one of the interpolation target coordinates, a target pixel group having the same degree of influence is selected from four target pixels at the interpolation target coordinates (n, m). select. Then, based on the pixel value of the selected target pixel group, the pixel value of the interpolation pixel at the interpolation target coordinates is calculated.

  That is, using the pixel values of the pixel of interest around the interpolation target coordinate and the reference pixels around the pixel of interest, the degree of influence of each pixel of interest at the interpolation target coordinate is calculated, and multiple attentions having the same degree of influence are calculated. From the pixel group, a set of target pixel groups (for example, a target pixel group having the highest influence degree among the target pixel groups having the same influence degree) is selected according to the degree of influence, and pixels of the selected target pixel group are selected. Based on the value (feature value), the pixel value of the interpolation pixel at the interpolation target coordinate is calculated. Each pixel of interest in the selected pixel group of interest has the same degree of influence, that is, a pixel of interest classified into the same attribute by the same attribute determination, and the pixel value of the interpolation pixel based on the pixel value (feature value) Therefore, no density or color blur occurs in the contour portion, and an interpolation result suitable for characters and text graphics can be obtained. In addition, since a set of target pixel groups is selected according to the degree of influence of each target pixel group around the interpolation target coordinates, smoothing processing can be performed on the contour shape of the drawing element, and the degree of influence can be calculated. Since the information to be used (for example, pixel values) is information of the input image (original image) before the interpolation process, a good smoothing result can be obtained regardless of the size of the high resolution ratio. Further, it is possible to obtain a high-resolution image free from jaggy and free from density blur and color mixing. Further, since the boundary of the drawing element is detected based on the difference in pixel value (feature amount) between the target pixels, pixel attribute information such as TAG information is not required for contour detection.

  In the present invention, for each target pixel, the presence / absence of the same attribute is determined for each pixel in the pixel block (reference pixel) including the target pixel and the plurality of target pixels. For example, when an interpolation pixel in a 2 × 2 pixel grid is obtained using a 2 × 2 pixel grid as a target pixel, a 4 × 4 pixel block around the 2 × 2 pixel grid is used as a reference pixel, The presence or absence of the same attribute as the reference pixel is determined. The presence / absence of the attribute can be determined based on, for example, whether the pixel values (feature values) are the same. The same attribute information assigned by associating the determination result of the presence or absence of the same attribute with each pixel in the pixel block (reference pixel) is extracted. For example, the same attribute information can be extracted for each pixel of interest as binarized information of “1” when the attribute is present and “0” when the attribute is absent. The extracted attribute information is interpolated at a predetermined interpolation magnification. In this case, for example, the attribute information can be multivalued with 0 and 100. Using the interpolated same attribute information (for example, 0 to 100), the degree of influence by surrounding pixels existing around the interpolation target coordinates is calculated. For example, the degree of influence can be calculated by applying a predetermined filter process to the interpolated same attribute information (for example, 0 to 100). In the filter processing, for example, the influence degree can be calculated using a moving average method by performing a matrix operation having a low-pass filter characteristic. Thereby, when calculating the influence degree, it is possible to increase the number of pixels to be referred to, and it is possible to obtain a smoother contour image in consideration of a wider range of information.

  In the present invention, the presence / absence of the attribute is determined for each target pixel depending on whether or not the pixel value of the target pixel matches each pixel in the pixel block (reference pixel). Note that the matching of pixel values includes partial or complete matching of pixel values (feature quantities). As a result, when the feature amount is different, it can be determined that there is no same attribute, and even when image regions of similar colors are adjacent to each other, each image region is correctly recognized as a different region, and Smoothing can be performed.

  In the present invention, the pixel value range in which the pixel value of each pixel in the pixel block (reference pixel) is provided for the pixel value of the target pixel is determined for each target pixel. Depending on whether it is included (whether or not they match). As a result, when an input image in which an image composed of pixels of the same color such as a character having a diagonal shape or curved edge shape is arranged on a gradation background is converted to high resolution, the diagonal edge portion or curved edge portion The occurrence of jaggies is suppressed, and the edge shape can be favorably retained.

  In the present invention, whether or not the attribute is present is determined by selecting a pixel value range based on the image attribute value of the target pixel for each target pixel, and whether or not the pixel value of the target pixel is included in the pixel value range. The presence / absence of the same attribute is determined according to whether the image attribute values are the same. Then, the pixel value of the interpolation pixel is calculated from the pixel value of the selected target pixel group according to the procedure selected according to the image attribute value of the target pixel group. That is, a different feature amount range is provided for each image attribute such as a character region, a vector graphics region, a natural image region such as a photograph, and the corresponding feature amount range is selected according to the image attribute value of each pixel. Perform attribute determination. For example, complete pixel values are matched for character areas, range is set to allow subtle color differences for vector graphics areas that include gradation, etc., and ranges are set for natural image areas such as photographs. The setting is maximized, the same attribute determination by color is not performed, and the same attribute is determined only when the image attribute values match.

  Then, the feature amount of the target pixel is copied as it is to the interpolation pixel associated with the target pixel whose image attribute value is the character region (character), and the target pixel whose image attribute value is the natural image region (natural image) is copied. For the associated interpolation pixel, the pixel value (feature amount) of each interpolation pixel is calculated by linear interpolation from the feature amounts of a plurality of target pixels determined to have the same attribute. This ensures exact color matching for characters, vector graphics with gradations have the same attribute determination within a certain range, and pixels with natural image attributes such as photographs are determined to have the same attribute. By maximizing the range of (i.e., determining that all have the same attribute), it is possible to perform suitable edge smoothing on the boundary between the photo area and the character area. That is, in the smoothing process using resolution conversion, when the same attribute is determined for the target pixel and the surrounding pixel group of the target pixel, and the pixel value of the interpolation pixel is determined based on the determination result, the selected same attribute layer By calculating the interpolated pixel value from the pixel values of multiple pixels of interest belonging to, the contour shape of the artificial image for an image that is a mixture of artificial images such as characters and vector graphics and natural images such as photographs Thus, a smooth image can be obtained in combination with a smoothing process for natural image density and a good image can be obtained.

  In the present invention, the presence / absence of the same attribute is determined for each target pixel, with each pixel in the pixel block (reference pixel) adjacent to the target pixel via the target pixel or a pixel that has already been determined to have the same attribute. Whether or not the same attribute is present is determined according to whether or not the same attribute is present (continuous). In this case, whether adjacent (continuous) or not can be performed in eight directions that are vertically and horizontally slanted. As a result, when there is a thin line sandwiched between two areas of the same color, such as a thin line drawn on the background color, when two areas sandwiching the thin line are recognized as one identical attribute area, The influence may be excessive, the fine line may become extremely thin, or the fine line may disappear. By considering continuity, even if the same color area is divided by thin lines, two areas of the same color can be recognized as separate areas, and the absolute value of the influence by each area becomes smaller, It is possible to prevent the fine line from becoming extremely thin or the fine line from disappearing.

  In the present invention, when selecting a pixel of interest corresponding to the interpolation target coordinates, the pixel of interest having the greatest influence is selected from the pixels of interest. As a result, the contour and the color are not blurred, and an interpolation result suitable for characters and text graphics can be obtained. In addition, smoothing processing can be performed on the contour shape of the drawing element, and a high-resolution image free from jaggy can be obtained.

  In the present invention, the calculated influence degree is compared with a first threshold value (for example, 50% of the maximum influence degree), and when the influence degree is smaller than the first threshold value, the target pixel closest to the interpolation target coordinate is determined. select. The interpolation processing basically performs smoothing processing by rounding edges, and has a feature that right-angled edge portions are also rounded. In particular, in an orthogonal part where three or more background colors are adjacent, such as when a straight line in the vertical direction and a straight line in the horizontal direction intersect on the background color, the part (edge) that should be a straight line is distorted. It will occur. The degree of influence is compared with the first threshold, and for the interpolation target coordinates with a small degree of influence, the pixel of interest closest to the interpolation target coordinate is determined by suppressing the selection of the pixel of interest with the largest degree of influence. By selecting a pixel of interest as a copy source and detecting a portion where drawing elements of three or more colors are orthogonal, such as the intersection of vertical and horizontal straight lines, and avoiding smoothing processing, linearity and orthogonality in the orthogonal portion Can keep.

  In the present invention, it is determined whether or not the pixel of interest belongs to an oblique line having a width of one pixel or a thin line that is an isolated point, and the influence degree at the interpolation target coordinates of the pixel of interest determined to belong to the thin line and the second Compare with the threshold. If the degree of influence at the interpolation target coordinates is larger than the second threshold value, the target pixel determined to belong to the thin line is selected in association with the interpolation target coordinates. If the degree of influence at the interpolation target coordinates is smaller than the second threshold, the pixel of interest having the highest degree of influence is selected in association with the interpolation target coordinates. Thereby, it is possible to prevent the oblique thin lines and the isolated points from disappearing by extracting the oblique thin lines and the isolated points and preferentially considering the degree of influence thereof.

  In the present invention, it is determined whether or not the pixel of interest belongs to an oblique line having a width of one pixel or a thin line that is an isolated point, and the influence degree at the interpolation target coordinates of the pixel of interest determined to belong to the thin line and the second Compare with the threshold. If the degree of influence at the interpolation target coordinates is larger than the second threshold value, the target pixel determined to belong to the thin line is selected in association with the interpolation target coordinates. When the influence degree at the interpolation target coordinates is smaller than the second threshold value, the calculated influence degree is compared with the first threshold value. When the influence level at the interpolation target coordinates is smaller than the first threshold value, the pixel of interest closest to the interpolation target coordinates is selected, and when the influence level at the interpolation target coordinates is equal to or greater than the first threshold value, the calculated attention level is the highest. Select a pixel. Thereby, it is possible to perform the smoothing process on the contour shape while maintaining the reproducibility of the oblique thin line and the isolated point and the linearity and orthogonality at the orthogonal portion.

  In the present invention, when it is determined that the pixel of interest and the pixel adjacent to the pixel of interest in the vertical or horizontal direction do not have the same attribute, it is determined that the pixel belongs to the thin line. For example, when the number of pixels that match the pixel value (feature value) of the target pixel and that are adjacent to the four vertical and horizontal directions (has continuity) is one pixel including the target pixel (that is, the target pixel itself) Only), it is determined that the target pixel belongs to a thin line. Thereby, even if it is a diagonal thin line of 1 pixel width or an isolated point, a line can be prevented from disappearing.

  In the present invention, the second threshold value is changed according to the number of pixels having the same attribute as the target pixel determined to belong to the thin line. If the fine line is long, the absolute value of the influence will be large, and if the fine line is short, the absolute value of the influence will be small. Therefore, if the second threshold value is used fixedly, the thin line after processing will have a long line with a wide width. , The width of the short line becomes narrower. By changing the second threshold according to the length of the line, the processed line width can be kept constant regardless of the length of the line.

  According to the present invention, it is possible to provide an image forming apparatus capable of obtaining an interpolation processing result suitable for characters and text graphics without causing density or color blur in the contour portion.

  According to the present invention, smoothing processing with reduced jaggies can be performed without causing blur due to interpolation processing in the contour portion.

1 is a block diagram illustrating a configuration of an image forming apparatus as a printer including an image processing apparatus according to the present invention. It is explanatory drawing which shows the positional relationship of the low resolution raster data which is input image data of a high resolution process part, and the high resolution raster data which are output image data. It is explanatory drawing which shows an example of input image data. It is explanatory drawing which shows an example of the attribute layer information. It is explanatory drawing which shows an example of the attribute layer information. It is explanatory drawing which shows the calculation method of an influence degree. It is explanatory drawing which shows the coordinate of each pixel of an input image, and an interpolation pixel. It is explanatory drawing which shows an example of the attribute layer information. It is explanatory drawing which shows an example of an influence degree. It is explanatory drawing which shows an example of the interpolation pixel by high resolution processing. It is explanatory drawing which shows an example of input image data. It is explanatory drawing which shows an example of the attribute layer information. It is explanatory drawing which shows an example of the attribute layer information. It is explanatory drawing which shows an example of the attribute layer information. It is explanatory drawing which shows an example of an influence degree. It is explanatory drawing which shows an example of the interpolation pixel by high resolution processing. It is explanatory drawing which shows the jaggy shape image which generate | occur | produces in the edge part of the character image on a gradation background image by high resolution processing. It is explanatory drawing which shows an example of the same attribute layer information at the time of using an effective determination range. It is explanatory drawing which shows an example of the same attribute layer information at the time of using an effective determination range. It is explanatory drawing which shows an example of the same attribute layer information at the time of using an effective determination range. It is explanatory drawing which shows an example of the influence degree at the time of using an effective determination range. It is explanatory drawing which shows an example of the interpolation pixel by the resolution increasing process at the time of using an effective determination range. It is explanatory drawing which shows an example of input image data. It is explanatory drawing which shows an example of the attribute layer information. It is explanatory drawing which shows an example of the attribute layer information. It is explanatory drawing which shows an example of the attribute layer information. It is explanatory drawing which shows an example of an influence degree. It is explanatory drawing which shows an example of the interpolation pixel by high resolution processing. It is explanatory drawing which shows an example of the attribute layer information at the time of using the effective determination range according to an image attribute value. It is explanatory drawing which shows an example of the attribute layer information at the time of using the effective determination range according to an image attribute value. It is explanatory drawing which shows an example of the attribute layer information at the time of using the effective determination range according to an image attribute value. It is explanatory drawing which shows an example of the influence degree at the time of using the effective determination range according to an image attribute value. It is explanatory drawing which shows an example of the interpolation pixel by the resolution increasing process at the time of using the effective determination range according to an image attribute value. It is explanatory drawing which shows an example of the interpolation pixel value calculation according to an image attribute value. It is explanatory drawing which shows the specific example of the interpolation pixel value calculation according to an image attribute value. It is explanatory drawing which shows the coordinate of each pixel of an input image which has an orthogonal part of 3 colors, and an interpolation pixel. It is explanatory drawing which shows an example of the same attribute layer information in the case of having an orthogonal part of three colors. It is explanatory drawing which shows an example of the influence degree in the case of having an orthogonal part of 3 colors. It is explanatory drawing which shows an example of the interpolation pixel by the high resolution process in the case of having an orthogonal part of 3 colors. It is explanatory drawing which shows the coordinate of each pixel and interpolation pixel of an input image which has an orthogonal part of 4 colors. It is explanatory drawing which shows an example of the same attribute layer information in the case of having an orthogonal part of 4 colors. It is explanatory drawing which shows an example of the influence degree in the case of having an orthogonal part of 4 colors. It is explanatory drawing which shows an example of the interpolation pixel by the high resolution process in the case of having an orthogonal part of 4 colors. It is explanatory drawing which shows an example of thin line layer information. It is explanatory drawing which shows an example of thin line layer information. It is explanatory drawing which shows the coordinate of each pixel of an input image, and an interpolation pixel. It is explanatory drawing which shows an example of the attribute layer information. It is explanatory drawing which shows an example of an influence degree. It is explanatory drawing which shows an example of thin line layer information. It is explanatory drawing which shows an example of the interpolation pixel by high resolution processing. An example of the threshold table of a thin line influence degree is shown. It is explanatory drawing which shows the coordinate of each pixel and interpolation pixel of the input image of 2 colors which consists of a character area and a vector graphics area. It is explanatory drawing which shows an example of the same attribute layer information when not using tag information. It is explanatory drawing which shows an example of the interpolation pixel by the resolution increasing process when not using tag information. It is explanatory drawing which shows an example of the attribute layer information at the time of using tag information. It is explanatory drawing which shows an example of the interpolation pixel by the resolution increasing process when not using tag information. It is explanatory drawing which shows a part of image after the high resolution process of this invention. It is a flowchart which shows the process sequence of the high resolution process of this invention. It is a flowchart which shows the process sequence of the high resolution process of this invention. It is explanatory drawing which shows a part of original image. It is explanatory drawing which shows a part of image after the process by the nearest neighbor interpolation as a prior art example. It is explanatory drawing which shows a part of image after the process by the linear interpolation as a prior art example. It is explanatory drawing which shows a part of image after the interpolation by the neighboring pixel comparison as a prior art example.

  Hereinafter, the present invention will be described with reference to the drawings illustrating embodiments. FIG. 1 is a block diagram showing a configuration of an image forming apparatus as a printer 2 including an image processing apparatus according to the present invention. As shown in FIG. 1, a printer 2 is connected to a computer 1 via a network or the like, and image data created on the computer 1 using application software 11 such as image editing software is paged by a printer driver 12. It is converted into a description language and transmitted to the printer 2.

  The printer 2 as an image forming apparatus includes an image processing unit 21 as an image processing apparatus, a print engine unit 22 as an image forming unit, and the like. The page description language data transmitted from the computer 1 is converted into engine output data by the image processing unit 21, and an image is formed on a paper medium or the like by the print engine unit 22.

  The image processing unit 21 includes a raster data generation unit 211, a high resolution processing unit 212, a halftone processing unit 213, and the like. The raster data generation unit 211 analyzes the page description language transmitted from the computer 1 and generates multivalued low-resolution raster data. Next, the high resolution processing unit 212 performs interpolation processing and smoothing processing on the low resolution raster data, generates multi-value high resolution raster data, and outputs it to the halftone processing unit 213. The halftone processing unit 213 converts the high resolution raster data into engine output data that matches the output format to the print engine unit 22.

  Since the page description language analysis and low-resolution raster data generation performed by the raster data generation unit 211 involve complicated arithmetic processing, a general-purpose CPU generally executes a program code that defines a predetermined processing procedure. The processing time varies greatly depending on the complexity of the page description language data. Furthermore, the processing time tends to increase in proportion to the square of the resolution of the low-resolution raster data to be output. Therefore, when generating high-resolution raster data from complex page description language data, it is not practical. Processing time may be required.

  On the other hand, the processing after raster data generation is a complex but routine process, so the processing must be performed within the required short time by using a DSP (Digital Signal Processor) or parallel processor, or by hardware. Is possible. For this reason, the raster data generation unit 211 generates low-resolution data to reduce the processing time that is variable and depends on the resolution, and the high-resolution processing unit 212 and the halftone processing unit 213 have a DSP, parallel processor, or The processing time can be shortened by hardware, and the processing time of the entire image processing unit 21 can be significantly shortened.

  Next, the resolution enhancement processing unit 212 will be described. The high resolution processing unit 212 includes the same attribute layer extraction unit 2121, a layer information interpolation unit 2122, a layer information filter unit 2123, a fine line layer extraction unit 2124, a copy source pixel selection unit 2125, a pixel feature amount copy unit 2126, and the like. Yes.

  The attribute layer extraction unit 2121 functions as an attribute determination unit and an extraction unit that extracts the attribute information, and from a low-resolution raster data as input image data, a target pixel and a pixel value (feature amount) around an interpolation target coordinate. ) Extract the same attribute layer information (same attribute information) indicating the same pixel region.

  A layer information interpolation unit 2122 serving as the same attribute information interpolation unit obtains an interpolation value (interpolation layer information) of the same attribute layer information at each interpolation target coordinate by a general interpolation process.

  The layer information filter unit 2123 serving as an influence degree calculating unit performs a low-pass filter process on the interpolation layer information, and outputs the result as an influence degree of the pixel of interest at each interpolation target coordinate.

  A thin line layer extraction unit 2124 as a thin line determination unit determines whether or not the target pixel is a thin line, and outputs the result as thin line layer information. Further, the fine line layer extraction unit 2124 has a function as changing means for changing the second threshold according to the number of pixels having the same attribute as the target pixel determined to belong to the fine line.

  The copy source pixel selection unit 2125 serving as a selection unit that selects a set of target pixel groups and an interpolation pixel value calculation unit associates each interpolation target coordinate with the degree of influence obtained by the layer information filter unit 2123 and the thin line layer information. A copy source pixel is selected and output as copy source pixel information. As a result, the pixel value (feature value) of the copy source pixel is copied as the pixel value of the interpolation pixel at each interpolation target coordinate.

  Further, the copy source pixel selection unit 2125 has a function as a first comparison unit that compares the influence level output from the layer information filter unit 2123 with a predetermined first threshold value, and the influence level and the predetermined second threshold value. It has a function as a second comparison means for comparing.

  The pixel feature amount copying unit 2126 generates high-resolution raster data by copying the pixel value (feature amount) of the copy source pixel selected in association with each interpolation target coordinate based on the copy source pixel information.

  Next, high resolution processing from low resolution raster data to high resolution raster data will be described. FIG. 2 is an explanatory diagram showing the positional relationship between the low-resolution raster data that is the input image data and the high-resolution raster data that is the output image data of the high-resolution processing unit 212. In FIG. 2, white circles indicate low resolution raster data as input image data, and black circles indicate high resolution raster data as output image data after interpolation processing. In the present embodiment, low resolution raster data of 4 × 4 pixels is subjected to interpolation of M times in the vertical direction (vertical direction) and N times in the horizontal direction (horizontal direction) to output high resolution raster data. The case will be described. Note that the number of pixels of raster data is not limited to the example of FIG. Further, dX and dY indicate deviation amounts (horizontal direction and vertical direction, respectively) between the pixel before interpolation and the nearest pixel after interpolation, and the interpolation magnifications M and N and deviation amounts dY and dX indicate a high amount. The coordinates of the interpolation pixel of the resolution raster data are determined.

  The high resolution processing in the present embodiment focuses on a 2 × 2 pixel grid in input image data (low resolution raster data), and the 2 × 2 pixels and surrounding 4 × 4 pixels (reference pixels). Based on the pixel information (for example, pixel value), a process for obtaining information (for example, pixel value) of the interpolation pixel at the interpolation target coordinates in the 2 × 2 pixel grid is a basic process. Then, an interpolated image of the entire input image is obtained by repeating the same processing while scanning the focused 2 × 2 pixel grid in the horizontal and vertical directions.

  The interpolation magnifications M and N are resolution ratios between the input image data and the output image data. The interpolation magnifications M and N can be processed even with non-integer values, but in that case, it is necessary to recalculate the relative coordinates of each interpolation pixel including the shift amounts dY and dX for each 2 × 2 pixel grid of interest. Therefore, the process becomes complicated. Therefore, the processing can be simplified by setting the interpolation magnifications M and N to integer values, and the shift amounts dY and dX can also be set to constant values. The example of FIG. 2 shows a case where M and N are integer values, and the shift amounts are dY = 1 / 2M and dX = 1 / 2N (assuming that the inter-pixel distance of the input image is 1).

  Next, a method for extracting the same attribute layer information as the same attribute information from the low resolution raster data will be described. FIG. 3 is an explanatory diagram showing an example of input image data, and FIGS. 4 and 5 are explanatory diagrams showing an example of the attribute layer information. In FIG. 3, it is assumed that 4 × 4 pixel input image data (low-resolution raster data) includes pixels of three colors, an A color pixel, a B color pixel, and a C color pixel. The feature amount of each pixel corresponds to, for example, a density value for a monochrome (monochrome) image, a density value for each color element such as RGB or CMYK for a color image, and further, an attribute (character, vector graphics) of each pixel. Additional information such as tag information indicating a photo area) can also be added. In the present embodiment, it is possible to determine that the pixels having the same density values of all the color elements have the same feature amount (the pixel values are the same). Note that it is possible to determine not only matching of all feature quantities but also partial matching. In addition, pixel feature amounts are represented by A, B, and C for convenience.

  4 (a), 4 (b), 5 (c), and 5 (d), respectively, obtain interpolation pixels (interpolation data) in the central 2 × 2 pixel grid shown in the example of FIG. For the purpose of 4 × 4 pixels of the input image data as reference pixels (pixel blocks), the four target pixels (1, 1), (1, 2), (2, 1), ( 2 and 2) shows a method of obtaining four sets of the same attribute layer information corresponding to each. The same attribute layer information is information indicating whether or not the target pixel and reference pixels existing in the vicinity thereof have the same attribute. For example, “1” is present if the same attribute exists, and “1” is present if there is no same attribute. By assigning “0”, it can be expressed as binarized data. Hereinafter, the method for obtaining the attribute layer information will be described in detail with reference to the example of FIG.

  FIG. 4A shows the same attribute layer information obtained by using the pixel (1, 1) as the target pixel in the 2 × 2 pixel grid. In FIG. 4A, the numerical value indicated in a circle indicating each pixel indicates the distance from the target pixel (1, 1), and the feature amount is the same in order from the pixel group having the smaller numerical value (distance). And determination of continuity of pixels. The determination of continuity is continuous when the pixel of interest or the same attribute pixel that has already been determined exists (adjacent) in the eight pixels (vertical, horizontal, and diagonal eight directions) around the determination target pixel. Judge that there is. By repeating these processes from the pixel group at the distance 1 to the pixel group at the distance 5, the determination is made for all the pixels. In FIGS. 4 and 5, pixels that are determined to have the same attribute as a result are indicated by black circles, and pixels that are determined not to have the same attribute are indicated by white circles.

  Next, a procedure for determining whether or not the attribute exists is described. First, for a pixel group whose distance from the target pixel (1, 1) is 1, as Step 1, the pixel (0, 1) is different from the target pixel (1, 1) and is continuous (adjacent). Therefore, it is determined that the attribute is not the same (no attribute is the same). Hereinafter, similarly, since the pixel (1, 0) is different in color from the target pixel (1, 1) and is continuous, it is determined to have the same attribute. Since the pixel (1, 2) is different from the target pixel (1, 1) and is continuous, the pixel (1, 2) is determined to have the same attribute. Since the pixel (2, 1) is different from the target pixel (1, 1) and is continuous, the pixel (2, 1) is determined to have the same attribute.

  Next, for the pixel group whose distance from the target pixel (1, 1) is 2, as Step 2, the pixel (0, 0) has the same color as the target pixel (1, 1) and is continuous (adjacent). Therefore, it is determined that the attribute is the same (has the same attribute). Similarly, since the pixel (0, 2) is different in color from the target pixel (1, 1) and is continuous, it is determined to have the same attribute. Since the pixel (2, 0) is different in color from the target pixel (1, 1) and is continuous, the pixel (2, 0) is determined to have the same attribute. Since the pixel (2, 2) has the same color as the pixel of interest (1, 1) and is continuous, the pixel (2, 2) is determined to have the same attribute.

  Next, with respect to the pixel group whose distance from the target pixel (1, 1) is 3, as Step 3, the pixel (1, 3) is different from the target pixel (1, 1) and has the same attribute. Since it is continuous with (2, 2), it is determined that the attribute is not the same. Similarly, the pixel (3, 1) has a different color from the target pixel (1, 1) and is continuous with the same attribute pixel (2, 2).

  Next, with respect to the pixel group whose distance from the target pixel (1, 1) is 4, as Step 4, the pixel (0, 3) is different in color from the target pixel (1, 1), and the continuous pixels are Because there is no attribute, it is determined that the attribute is not the same. Since the pixel (2, 3) is different in color from the pixel of interest (1, 1) and is continuous with the same attribute pixel (2, 2), it is determined that the pixel (2, 3) has the same attribute. The pixel (3, 0) is different in color from the target pixel (1, 1) and has no continuous pixel, so it is determined to have the same attribute. Since the pixel (3, 2) is different in color from the target pixel (1, 1) and is continuous with the same attribute pixel (2, 2), the pixel (3, 2) is determined to have the same attribute.

  Next, with respect to the pixel group whose distance from the target pixel (1, 1) is 5, as Step 5, the pixel (3, 3) is the same color as the target pixel (1, 1) and has the same attribute. Since it is continuous with (2, 2), it is determined that the attribute is the same.

  This completes the determination of the same attribute for all pixels. The region information with the same attribute pixel as 1 and the non-same attribute pixel as 0 according to the determination result is output as the same attribute layer information of the pixel of interest (1, 1). The same processing is performed for the target pixel (1, 2), (2, 1), (2, 2) in the same manner, so that four pieces of attribute layer information corresponding to the four target pixels can be obtained. .

  For the same attribute layer information obtained as described above, the layer information interpolation unit 2122 performs multi-value conversion and interpolation processing. Since the attribute layer information is binary information of “1” and “0”, the influence degree can be expressed in multiple values by converting this information into multiple values. The position of the interpolation target coordinates conforms to the coordinate position of the high resolution raster data shown in FIG. The layer information interpolation unit 2122 calculates the interpolation value of the same attribute layer at the interpolation target coordinates (the coordinates of the high resolution raster data), and outputs it as interpolation layer information. In this embodiment, multi-value data is set to 0 and 100, and an interpolation value that is interpolation layer information is also output as a value within a range of 0 to 100. Note that multi-value data is not limited to 0 and 100, and the range of interpolation values is not limited to 0 to 100.

  As the interpolation method for obtaining the interpolation layer information, it is necessary to have a significant difference in output value depending on the distance between the pixel of the original image and the interpolation pixel for subsequent processing. Can not do it. In the present embodiment, linear interpolation is used as the simplest interpolation method that satisfies the above conditions among general interpolation methods, but other nonlinear interpolation can also be used.

  FIG. 6 is an explanatory diagram showing a method for calculating the influence degree. In FIG. 6, white circles in the 4 × 4 pixel area indicate interpolation layer information that is input data to the layer information filter unit 2123, and black circles indicate output data from the layer information filter unit 2123. Indicates a certain degree of influence. As shown in FIG. 6, the influence degree is obtained for each of the four target pixels of the 2 × 2 pixel grid corresponding to each interpolation target coordinate in the 2 × 2 pixel grid.

Specifically, as shown in FIG. 6, a matrix operation having a low-pass filter characteristic is performed on input data within a predetermined range (a range indicated by a broken line in FIG. 6) centering on the interpolation target coordinates for obtaining the influence degree. Thus, individual output data can be obtained. In the present embodiment, the moving average method, which is one of the simplest low-pass filters, is used, and the matrix calculation coefficients are all 1 / (total number of pixels within a predetermined range). The output value of this low-pass filter is output as the degree of influence by the pixel of interest at each interpolation target coordinate.

  Although accuracy is reduced, the number of pixels belonging to the same color in a predetermined block (for example, 3 × 3 pixels around each target pixel) is obtained for each target pixel, and the interpolation target coordinates are calculated from the target pixel as the value. The degree of influence of the pixel of interest on the interpolation target coordinates may be obtained by multiplying the distance up to.

  Next, basic processing for high resolution processing will be described. FIG. 7 is an explanatory diagram showing coordinates of each pixel and interpolation pixel of the input image, FIG. 8 is an explanatory diagram showing an example of the same attribute layer information, and FIG. 9 is an explanatory diagram showing an example of the influence degree, FIG. 10 is an explanatory diagram showing an example of an interpolated pixel obtained by the high resolution processing. In FIG. 7, white circles indicate the coordinates of each pixel of the input image data, and black dots indicate the coordinates of the interpolation pixel after the high resolution processing that is the output image data, and indicate the relationship between the positions. A and B in white circles indicate pixel colors (features), respectively. In addition, each pixel of the center 2 × 2 pixel lattice is the target pixel (1, 1), (1, 2), (2, 1), (2, 2).

  8A, FIG. 8B, FIG. 8C, and FIG. 8D are the target pixels (1, 1), (1, 2), (2) indicated by thick circles in FIG. 1), (2, 2) corresponding attribute layer information (1, 1), attribute layer information (1, 2), attribute layer information (2, 1), attribute layer information (2, 2) ). Numerical values in circles indicate the presence or absence of the same attribute, “1” indicates the same attribute (with the same attribute), and “0” indicates the non-same attribute (without the same attribute).

  9 (a), 9 (b), 9 (c), and 9 (d) show the target pixel (1, 1), (1, 2), (2, 1), (2, 2). Indicates the degree of influence. In FIG. 9, the numerical value in the circle indicating the interpolation target coordinates represents the degree of influence (for example, a range of 0 to 100) by the target pixel at each interpolation target coordinate. The interpolation target coordinates are assumed to be 4 × 4 pixel coordinates (4 × 4 pixel interpolation pixels) for convenience in an area surrounded by four pixels of interest in a 2 × 2 pixel grid.

  FIG. 10 shows the result of selecting the pixel of interest having the greatest influence by comparing the degrees of influence of the four pixels of interest at the 4 × 4 pixel interpolation target coordinates. For example, as shown in FIG. 9, since the influence degree in the upper left interpolation target coordinates is 87 for the target pixel A and 13 for the target pixel B, the target pixel A is selected. Similarly, the degree of influence at the lower right interpolation target coordinate is 41 for the target pixel A and 59 for the target pixel B, and therefore the target pixel B is selected. The same applies to other interpolation target coordinates.

  The pixel having the maximum influence level shown in FIG. 10 is set as a copy source pixel, and the pixel feature value copying unit 2126 copies the feature value of the target pixel selected from the copy source pixel (selected target pixel) to the interpolation target coordinates. High resolution raster data can be obtained. In the present embodiment, 4 × 4 pixels of the input image data are referred to as the reference pixels, but the reference pixels as the pixel blocks are not limited to this, and the reference pixels By further widening the range, it is possible to make the contour of the output image smoother. A smoothing result having an arbitrary smoothness can be obtained by designating a reference range according to the purpose.

  In the description using FIGS. 2 to 10 described above, it is determined that the pixels having the same density value of all the color elements have the same feature amount (the pixel values are the same). However, the present invention is not limited to this. Instead of providing a predetermined range for the feature amount, it can be determined that the feature amount is matched within the range. That is, for each target pixel, whether or not the pixel value of each pixel in the reference pixel (pixel block) is included in the pixel value range (feature amount range) provided for the pixel value of the target pixel (match) The presence or absence of the same attribute can be determined according to whether or not to do so. Hereinafter, this point will be described.

  11 is an explanatory diagram illustrating an example of input image data, FIGS. 12 and 13 are explanatory diagrams illustrating an example of the same attribute layer information, and FIG. 14 is an explanatory diagram illustrating an example of the same attribute layer information. FIG. 15 is an explanatory diagram illustrating an example of the degree of influence, and FIG. 16 is an explanatory diagram illustrating an example of an interpolation pixel obtained by the high resolution processing.

  As shown in FIG. 11, a 4 × 4 input image in which the feature amount of each pixel of the input image data is an RGB value will be described as an example. In this example, R, G, and B each take a value of 0 to 255, and the RGB value of each coordinate is as follows. (1) Pixels whose coordinates are (2, 1) and (3, 2) are (R, G, B) = (0, 255, 0). (2) The pixels whose coordinates are (1, 0), (2, 0), (3, 1) are (R, G, B) = (12, 255, 12). (3) Pixels with coordinates (0, 0) and (3, 0) are (R, G, B) = (24, 255, 24). (4) The pixels of other coordinates are (R, G, B) = (0, 0, 0). Hereinafter, for convenience, the feature quantities (1) to (4) will be referred to as A, A ′, A ″, and B, respectively.

  That is, the input image shown in FIG. 11 is an example in which black single data such as characters are arranged on a gradation (continuous tone) background in which pixel values gradually change.

  For the input image shown in FIG. 11, as described above, when the same attribute layer is determined on the condition that all feature quantities match, the target pixels (1, 1), (1, 2), (2 The same attribute layer information for 1), (2, 2) is as shown in FIGS. 12 and 13, as in FIGS. 4 and 5, the characters in each circle indicating an input pixel indicate the distance from each pixel of interest, and the black circle indicates each of the above-described pixels. A pixel that is determined to have the same attribute with respect to the target pixel, and a white circle indicates a pixel that is determined not to have the same attribute. In the notation of FIG. 14, as in FIG. 8, the characters in the white circles indicating the input pixels have the same attribute for each pixel of interest (“1”: same attribute, “0”: non-identical) Attribute). Output pixels (interpolation pixels) are indicated by black circles.

  Further, the degree of influence of each pixel of interest on each interpolation pixel (output pixel) in the 2 × 2 grid composed of each pixel of interest is as shown in FIG. 15, as in FIG. 9, the interpolation pixel configuration is a 4 × 4 pixel configuration, and the degree of influence of each pixel of interest is expressed within a range of 0 to 100. Further, FIG. 16 shows an output image (interpolated pixel distribution) obtained by selecting the pixel value of interest having the greatest influence from the results of FIGS. 15 (a) to 15 (d).

  As a result, when the same attribute layer is determined on the condition that all feature quantities match, the surrounding pixels of the target pixel (2, 1) shown in FIG. 11 have the same feature quantity A as (2, 1). Including the pixel (3, 2), there are a total of 6 pixels (1 pixel (3, 2) of the feature amount A and 3 pixels (1) of the feature amount A 'having the feature amount very close to the pixel (2, 1). , 0), (2, 0), (3, 1), and two pixels (0, 0), (3, 0) of feature amount A ″), the same attribute is determined. Since there is only one pixel (3, 2), the feature amount is A with respect to the degree of influence of the other three pixels of interest on the interpolation pixel of the 2 × 2 grid whose feature amount is B. The degree of influence of a certain pixel of interest (2, 1) on the interpolation pixel is reduced, and the feature amount of all the interpolation pixels is B.

  As described above, in the case of the same attribute layer extraction process in which it is determined that the same attribute layer is obtained by matching all the feature amounts, the same color pixel is formed on the gradation background in which the pixel value gradually changes as in the example illustrated in FIG. When an image (for example, a character image) is arranged and the pixel configuration of a 2 × 2 grid in the vicinity where high-resolution raster data is generated is only two colors (3 to 4 background colors as described later) If the pixel distribution is not adjacent to each other), the feature amount of the high-resolution raster data after the resolution conversion at that location tends to be deflected to the feature amount of the same color pixel.

  FIG. 17 is an explanatory view showing a jaggy shape image generated at the edge portion of the character image on the gradation background image by the high resolution processing. As described above, when an input image in which characters having an oblique shape or a curved edge shape are arranged on a gradation background is subjected to high resolution conversion by the above method, the original oblique shape or curved shape must be retained. As shown in FIG. 16, the rectangular image in which the same color image side is emphasized is more likely to be output at the location. For this reason, as shown in FIG. 17, there are cases where jaggy is likely to occur at the oblique edge portion and the curved edge portion of the character after resolution conversion.

  Therefore, by providing a predetermined range for the feature amount determined to be “same attribute” with respect to the feature amount of the target pixel, and extracting all surrounding pixels corresponding to the feature amount as the same attribute pixel, Occurrence can be suppressed. Hereinafter, this point will be described.

The feature amount, that is, the RGB value of each pixel of interest (1, 1), (1, 2), (2, 1), (2, 2) in FIG. 11 is defined as (R _CT , G _CT , B _CT ). For this RGB value, the upper and lower thresholds for setting the feature value range are R _UP and R _DN for the R value, G _UP and G _DN for the G value, respectively. For B value, B _UP and B _DN are used. Then, all the surrounding pixels having RGB values (R _X , G _X , B _X ) corresponding to all of the expressions (1) to (3) are determined as the same attribute pixels with respect to the target pixel, and the subsequent processing is performed. . The range represented by Expression (1) to Expression (3) is an effective determination range for each of the RGB values. Also, appropriate values can be set for the upper and lower thresholds.

For example, when R _up = R _DN = 32, G _up = G _DN = 16, and B _up = B _DN = 32, the feature amount A (R, G, B) of the target pixel (2, 1) in FIG. = (0, 255, 0) The feature amount range (R _X , G _X , B _X ) determined as “same attribute” is 0 ≦ R _X ≦ 32, 239 ≦ G _X ≦ 255, 0 ≦ B _X ≦ 32.

Similarly, when R _up = R _DN = 32, G _up = G _DN = 16, and B _up = B _DN = 32, the target pixel (1, 1), (1, 2), ( The feature amount range (R _X , G _X , B _X ) determined as “same attribute” for the feature amount B (R, G, B) = (0, 0, 0) of 2 and 2) is 0 a _{≦ R X ≦ 32,0 ≦ G X} ≦ 16,0 ≦ B X ≦ 32.

  18 and 19 are explanatory diagrams showing an example of the same attribute layer information when the valid determination range is used, and FIG. 20 is an explanatory diagram showing an example of the same attribute layer information when the valid determination range is used. FIG. 21 is an explanatory diagram illustrating an example of the degree of influence when the effective determination range is used, and FIG. 22 is an explanatory diagram illustrating an example of the interpolation pixel by the resolution enhancement process when the effective determination range is used.

  18 to 20 show respective target pixels (1, 1), (1, 2), (2, 1) when the same attribute pixel determination method using the above numerical example is applied to the input pixel shown in FIG. ), (2, 2) indicates the same attribute layer information. The notations in FIGS. 18 and 19 are the same as the notations in FIGS. 12 and 13, and the notations in FIG. 20 are the same as the notations in FIG.

  As shown in FIGS. 18 to 20, when a feature amount of a pixel determined to have the same attribute with respect to the target pixel is given a range, it is determined that the pixel has the same attribute when all the feature amounts match (FIGS. 12 to 20). 14), the five pixels (features A ′ (1, 0), (2, 0), (3, 1)) that were not determined to have the same attribute as the target pixel (2, 1). , Feature quantity A ″ (0, 0), (3, 0)) is determined as the same attribute pixel of the target pixel (2, 1). Thereby, each interpolation in the 2 × 2 grid The influence degree and output image (interpolated pixel distribution) of each pixel of interest of the pixel (output pixel) are as shown in FIGS. 21 and 22, respectively.

  That is, the influence degree of the target pixel (2, 1) (feature amount A) on the target pixel (1, 1), (1, 2), (2, 2) whose feature amount is B is relatively large, As a result, compared to FIG. 16, that is, compared to the case where it is determined that the same attribute pixel is obtained by matching all the feature amounts, the feature amounts of the three interpolation pixels located near the target pixel (2, 1) are A (see FIG. 22). ), The occurrence of jaggy at the oblique edge portion and the curved edge portion of the image composed of the same color pixels after the high resolution conversion is suppressed.

The parameters R _up , R _DN , G _up , G _DN , B _up, and B _DN that determine the range to be determined as the same attribute pixel are input to the same attribute layer extraction unit 2121 shown in FIG. 1 as changeable parameters. The configuration is as follows. The setting and input method may be input by the user from the operation panel of the image forming apparatus, or a block for analyzing the image information of the raster image obtained by the raster data generation unit 211 shown in FIG. 1 is provided. A method of calculating and inputting each parameter from the result may be used.

  In the description using FIG. 11 to FIG. 22 described above, when the same attribute is determined, a predetermined range (effective determination range) is provided for the feature amount, so that the feature amount is consistent within the range. The feature amount range common to all the regions is used, but the feature amount range may be switched according to an image attribute previously assigned to each pixel. Note that image attributes can be distinguished by, for example, image attribute values.

  That is, a different feature amount range is provided for each image attribute such as a character region, a vector graphics region, a natural image region such as a photograph, and the corresponding feature amount range is selected according to the image attribute value of each pixel. Perform attribute determination. For example, complete pixel values are matched for character areas, range is set to allow subtle color differences for vector graphics areas that include gradation, etc., and ranges are set for natural image areas such as photographs. The setting is maximized, the same attribute determination by color is not performed, and the same attribute is determined only when the image attribute values match. Hereinafter, this point will be described.

  In the following description, it will be explained that jaggies are likely to occur when the same attribute layer is determined when all feature quantities match, and then, an effective determination range according to an image attribute value given in advance for each pixel. It will be described that by using this to determine the same attribute layer, suitable edge smoothing can be performed for the boundary between the photo area and the character area.

  First, a case will be described in which it is determined that the same attribute layer is obtained when all of the feature amounts match. FIG. 23 is an explanatory diagram illustrating an example of input image data, FIGS. 24 and 25 are explanatory diagrams illustrating an example of the same attribute layer information, and FIG. 26 is an explanatory diagram illustrating an example of the same attribute layer information. FIG. 27 is an explanatory diagram illustrating an example of the degree of influence, and FIG. 28 is an explanatory diagram illustrating an example of an interpolated pixel obtained by the high resolution processing.

As shown in FIG. 23, a 4 × 4 input image in which an RGB value and an image attribute value T are given as the feature amount of each pixel of the input image data will be described below as an example. In this example, R, G, and B each have a value of 0 to 255, and the image attribute value T has a value of 0 (character) and 1 (natural image). The RGBT values of each coordinate are as shown in FIG. 23, and the coordinates are (0, 0), (1, 0), (2, 0), (2, 1), (3, 0), (3, The image attribute values T of the pixels 1), (3, 2), and (3, 3) are 0 (characters) and all have the same RGB value. Hereinafter, for the sake of convenience, the feature amount of each pixel is denoted as A. In addition, the image attribute value T of pixels other than the above-described pixels is 1 (natural image), and the RGB values have different values for all pixels. Hereinafter, for the sake of convenience, the feature quantities of the pixels at these other coordinates are referred to as B _{01 to} B ₂₃ in association with the coordinates.

  That is, the input image shown in FIG. 23 shows a case where monochromatic characters are arranged on the lower surface of a natural image such as a photograph whose pixel values change randomly for each coordinate. This is an example of an image in which artificial images such as characters and vector graphics are mixed.

  When the same attribute layer is determined for the input image shown in FIG. 23 on the condition that all feature quantities match, the target pixels (1, 1), (1, 2), (2, 1), ( The same attribute layer information for 2 and 2) is as shown in FIGS. 24 and 25, as in FIG. 4 and FIG. 5, the characters in each circle indicating the input pixel indicate the distance from each pixel of interest, and the black circle indicates each of the above. A pixel that is determined to have the same attribute with respect to the target pixel, and a white circle indicates a pixel that is determined not to have the same attribute. In the notation of FIG. 26, as in FIG. 8, the characters in the white circles indicating the input pixels have the same attribute for each pixel of interest (“1”: same attribute, “0”: non-identical). Attribute). Output pixels (interpolation pixels) are indicated by black circles.

  Further, the degree of influence of each pixel of interest on each interpolation pixel (output pixel) in the 2 × 2 grid composed of each pixel of interest is as shown in FIG. 27, as in FIG. 9, the interpolation pixel configuration is a 4 × 4 pixel configuration, and the degree of influence of each pixel of interest is expressed within a range of 0 to 100. Furthermore, FIG. 28 shows an output image (interpolated pixel distribution) obtained by selecting the pixel value of interest having the greatest influence from the results of FIGS. 27 (a) to 27 (d).

As a result, when the same attribute layer is determined on the condition that all feature quantities match, the surrounding pixels of the target pixel (1, 1), (1, 2), (2, 2) shown in FIG. Since there are 8 pixels of the feature amounts B _{01 to} B ₂₃ that are considered to be included in the same drawing object as a natural image area, the pixel values are all different, so the number of pixels determined to have the same attribute is (1, 1), (1,2), (2,2) are one pixel. Further, regarding the peripheral pixels of the target pixel (2, 1), the pixels having the same feature amount A are 8 pixels. As a result, in most of the interpolation pixels, the degree of influence of the target pixel (1, 1), (1, 2), (2, 2) is smaller than the degree of influence of the target pixel (2, 1). feature quantity becomes a, only, only one pixel nearest the pixel of interest (1,2) is the feature amount B _12.

  As described above, in the case of the same attribute layer extraction process in which it is determined that the same attribute layer is obtained by matching all the feature amounts, a natural value such as a photograph whose pixel value changes randomly for each coordinate as in the example shown in FIG. An image composed of pixels of the same color (for example, a character image) is arranged, and the feature amount of the high-resolution raster data after resolution conversion at that portion tends to be deflected to the feature amount of the pixel of the same color. As in the case of the jaggy shape image generated at the edge portion of the character image on the gradation background image shown, jaggy is likely to occur at the oblique edge portion and the curved edge portion of the character after resolution conversion.

  Next, a case where the attribute layer is determined using an effective determination range corresponding to an image attribute value given in advance for each pixel will be described. In this case, the range of the feature amount determined to be “same attribute” with respect to the feature amount of the target pixel is switched by the image attribute value T assigned to each pixel, the image attribute value falls within the range. By extracting all the surrounding pixels having the same T as the same attribute pixels, the above-described jaggy can be suppressed. Hereinafter, this point will be described.

  In the example of FIG. 23, when the image attribute value T of the target pixel is 0 (character), the comparison of RGB pixel values is determined as the same attribute by strict matching, and the image attribute value T is 1 (natural image). In the case of), the pixel value comparison range is set to the maximum width (0 to 255), and the determination based on the color is not substantially performed, and the same attribute is determined only by the matching of the image attribute values.

  29 and 30 are explanatory diagrams showing an example of the same attribute layer information when an effective determination range according to an image attribute value is used, and FIG. 31 shows a case where an effective determination range according to an image attribute value is used. FIG. 32 is an explanatory diagram showing an example of the attribute layer information, FIG. 32 is an explanatory diagram showing an example of the degree of influence when an effective determination range according to the image attribute value is used, and FIG. 33 is a diagram according to the image attribute value It is explanatory drawing which shows an example of the interpolation pixel by the resolution increasing process at the time of using an effective determination range.

  29 to 31 show the respective pixel of interest (1, 1), (1, 2), (2,...) When the same attribute pixel determination method using the above numerical example is applied to the input pixel shown in FIG. The same attribute layer information for 1), (2, 2) is shown. Each notation in FIGS. 29 and 30 is the same as each notation in FIGS. 12 and 13, and each notation in FIG. 31 is the same as each notation in FIG. 14.

As shown in FIGS. 29 to 31, the target pixels (1, 1), (1, 2), (2, 2) whose image attribute value T is 1 (natural image) are pixels that are determined to have the same attribute. Since the feature amount is in the maximum range and is determined to be the same attribute pixel only by matching the image attribute value T, all the pixels of the feature amounts B _{02 to} B ₂₃ that have not been determined to have the same attribute in the examples of FIGS. The same attribute is determined, and the number of pixels is 8.

  On the other hand, for the target pixel (2, 1) having an image attribute value T of 0 (character), the feature amount is determined with the exact match of the feature amount, so that the feature amount is the same as in the examples of FIGS. All the pixels of A are determined to have the same attribute, and the number of pixels is 8. As a result, the interpolated pixel in which the influence degree of the target pixel (2, 1) is larger than the influence degree of (1, 1), (1, 2), (2, 2) is the pixel of interest (2, 1). There are only three neighboring pixels, and the remaining pixels have a greater influence due to (1, 1), (1, 2), (2, 2).

As a result, the feature amount given to the interpolation pixel is as shown in FIG. That is, the three pixels in the vicinity of the target pixel (2, 1) are given the feature amount A because the degree of influence by the target pixel (2, 1) is large. For the other interpolated pixels, the degree of influence by the target pixel (1, 1), (1, 2), (2, 2) having the same attribute is equal and maximum, but the feature amount B _{11 of} each target pixel. , B ₁₂ , and B ₂₂ each have an individual value, and therefore feature values based on these feature values are calculated for each interpolation pixel. Here, a symbol b representing the feature amount of the pixel group is shown.

Here, the boundary between the pixel group of the feature amount A and the pixel group of the feature amount b is a boundary when the natural image regions from the feature amount B ₁₂ to the feature amount B ₂₃ are the same region, and the smoothing result of the object contour As a boundary, the occurrence of jaggies that should be obtained is suppressed.

  FIG. 34 is an explanatory diagram showing an example of calculating an interpolated pixel value according to an image attribute value. FIG. 34 shows the details of the feature amount of each interpolation pixel, and shows an example in which the calculation method of the feature amount of the interpolation pixel is switched according to the image attribute value T of the corresponding target pixel. In this example, the feature amount of the target pixel is copied as it is to the interpolation pixel associated with the target pixel whose image attribute value T is 0 (character), and the target pixel whose image attribute value T is 1 (natural image). For the interpolation pixel associated with, the feature amount of each interpolation pixel is obtained by linear interpolation from the feature amounts of a plurality of target pixels determined to have the same attribute.

  That is, the interpolation pixel <2, 0>, <3, 0>, <3, 1> is associated with the target pixel (2, 1) as a result of the influence comparison, and the image attribute value T is 0 (character ), The feature amount A is copied as it is. The other interpolation pixels are associated with the target pixel (1, 1), (1, 2), (2, 2) having the same attribute, and the image attribute value T is 1 (natural image). B00 to b33 are given as feature amounts obtained by linear interpolation based on the feature amounts of the target pixel (1, 1), (1, 2), (2, 2). As linear interpolation, it is normal to calculate an interpolation value based on four surrounding pixels, but in this embodiment, an interpolation value is calculated based only on the corresponding pixel of interest having the same attribute. Therefore, the pixels that are not associated with the same attribute among the surrounding four pixels are calculated by giving a temporary pixel value.

FIG. 35 is an explanatory diagram showing a specific example of interpolation pixel value calculation according to the image attribute value. As shown in FIG. 35, in this example, the pixel (2, 1) that was the feature amount A in FIG. 34 is replaced with a white pixel (feature amount W). As a result, the pixel value of each interpolation pixel is calculated by linear interpolation based on the feature amounts B ₁₁ , B ₁₂ , W, and B ₂₂ of the four target pixels. As a result, among the interpolated pixels, the color change between the pixels in the natural image region can be made a smooth change suitable for the natural image.

  By the method described above, contour smoothing between objects is performed on images where artificial drawing objects such as characters and vector graphics and natural image objects such as photographs are mixed, and the color change in each object is An interpolated image can be obtained in a state suitable for each object. That is, for example, exact color matching is performed for characters, the same attribute is determined for a certain range in vector graphics that include gradation, and the same attribute is used for pixels with natural image attributes such as photographs. By maximizing the determination range (that is, determining that all have the same attribute), it is possible to perform suitable edge smoothing on the boundary between the photo area and the character area.

  As described above, in the smoothing process using resolution conversion, the same attribute determination is performed on the target pixel and the surrounding pixel group of the target pixel, and the pixel value of the interpolation pixel is determined based on the determination result. By calculating the interpolated pixel value from the pixel values of multiple pixels of interest belonging to the attribute layer, an artificial image such as a character or vector graphics and a natural image such as a photograph are mixed. It is possible to obtain a good image by performing both the smoothing process for the contour shape and the smoothing process for the density of the natural image.

  In this embodiment, as an example of the interpolation pixel calculation method, copying of the target pixel value and linear interpolation based on the target pixel value of the same attribute are used. Alternatively, other general interpolation methods such as nonlinear interpolation may be applied. Further, the pixels referred to when calculating the interpolation pixel may include not only the target pixel but also surrounding pixels determined to have the same attribute.

  In this embodiment, the image attribute value of each pixel is assigned in advance. However, a separate region determination unit may be provided to determine the image attribute value. As a result, even for an image input from a scanner or a digital camera, it is possible to obtain a desired result by obtaining an image attribute value and processing a character area and a natural image area such as a photograph separately.

  Next, processing for maintaining linearity and orthogonality in orthogonal portions where three or more background colors are adjacent, such as when a vertical line and a horizontal line intersect on the background color, will be described. To do. FIG. 36 is an explanatory diagram illustrating coordinates of each pixel and interpolation pixel of an input image having three color orthogonal portions, and FIG. 37 is an explanatory diagram illustrating an example of the same attribute layer information in the case of having three color orthogonal portions. FIG. 38 is an explanatory diagram showing an example of the degree of influence in the case of having three color orthogonal portions, and FIG. 39 is an explanatory diagram showing an example of an interpolation pixel by high resolution processing in the case of having three color orthogonal portions. It is. In FIG. 36, white circles indicate the coordinates of each pixel of the input image data, and black dots indicate the coordinates of the interpolation pixel after the resolution enhancement processing that is the output image data, and indicate the relationship between the positions. A, B, and C in white circles indicate the color (feature amount) of each pixel. The example of FIG. 36 shows a state where the regions represented by colors A, B, and C are separated by a vertical straight line and a horizontal straight line. In addition, each pixel of the center 2 × 2 pixel lattice is the target pixel (1, 1), (1, 2), (2, 1), (2, 2).

  37 (a), 37 (b), 37 (c), and 37 (d) are pixels of interest (1, 1), (1, 2), (2) indicated by thick circles in FIG. 1), (2, 2) corresponding attribute layer information (1, 1), attribute layer information (1, 2), attribute layer information (2, 1), attribute layer information (2, 2) ). Numerical values in circles indicate the presence or absence of the same attribute, “1” indicates the same attribute (with the same attribute), and “0” indicates the non-same attribute (without the same attribute).

  38 (a), 38 (b), 38 (c), and 38 (d) show the target pixel (1, 1), (1, 2), (2, 1), (2, 2). Indicates the degree of influence. In FIG. 38, numerical values in circles indicating interpolation target coordinates represent the degree of influence (for example, a range of 0 to 100) at each interpolation target coordinate. The interpolation target coordinates are assumed to be 4 × 4 pixel coordinates (4 × 4 pixel interpolation pixels) for convenience in an area surrounded by four pixels of interest in a 2 × 2 pixel grid.

  The interpolation processing basically performs smoothing processing by rounding edges, and has a feature that right-angled edge portions are also rounded. In particular, as shown in FIG. 36, in the orthogonal portion where three or more background colors are adjacent, such as when a vertical straight line intersects a horizontal straight line on the background color, it should be a straight line. The part (edge) is distorted.

  Therefore, as shown in FIG. 39 (a), when comparing the degree of influence of the four target pixels at each interpolation target coordinate, when the degree of influence is interpolated with a first threshold value (for example, a value of 0 to 100), the upper limit 50% value of value 100), and for interpolation target coordinates whose maximum influence is less than 50% value, it is possible to suppress the selection of the target pixel having the greatest influence and to select the interpolation target coordinates. The nearest pixel of interest is selected as the pixel of interest for copying. In the example of FIG. 39A, the pixel value of the interpolation pixel of the interpolation target coordinate indicated by “X” in a circle is copied, and the pixel value of the target pixel closest to the interpolation target coordinate is copied. Thereby, the linearity and orthogonality in the orthogonal part can be maintained by detecting a part where drawing elements of three or more colors are orthogonal to each other, such as an intersection of vertical and horizontal straight lines, and avoiding the smoothing process.

  On the other hand, FIG. 39C shows the nearest pixels (pixels A, B, and C) at each interpolation target coordinate, and the interpolation target coordinates indicated by “X” in a circle in FIG. 39A. The last selected pixel shown in FIG. 39B can be obtained by replacing this pixel with the nearest pixel in FIG. By copying the pixel feature amount using the final selected pixel as a copy source pixel, high-resolution raster data that maintains the linearity of the orthogonal portion can be obtained.

  Note that if the process of selecting the target pixel closest to the interpolation target coordinates by comparing the degree of influence with the threshold value as described above is not performed, in the final selected pixel shown in FIG. Among the 4 × 4 interpolation pixels, the pixel values of the interpolation pixels in the lower three rows are all C. As described above, the influence is compared with the threshold value and the processing for selecting the pixel of interest closest to the interpolation target coordinates is performed, so that the orthogonal portion divided by the three colors of pixels A, B, and C becomes the pixel C. It is possible to further prevent the edges from being rounded.

  That is, as shown in FIG. 36, in the portion where the three colors (A color, B color, and C color) are orthogonal, among the 4 × 4 input pixels, each of the A color and B color that is a right angle region is 4 pixels, The C color in the straight line area is 8 pixels. In the example of FIG. 36, the two colors A and B are symmetric with respect to the center line in the vertical direction (longitudinal direction), so that the influences thereof are also symmetric and the boundary after smoothing processing is performed. Also coincides with the vertical centerline. In order to maintain the orthogonality of the three colors, the boundary between the A color region and the B color region and the C color region needs to coincide with the horizontal (horizontal) center line. Each of the B and B color areas has a smaller number of pixels than the C color area. Therefore, when the degree of influence is compared, the degree of influence of the C color area is large and the boundary line is biased upward from the center line.

  On the other hand, if attention is paid only to the C color region, the number of C color pixels corresponds to 50% of the entire input pixels, and therefore the boundary line having an influence degree of 50% coincides with the center line in the horizontal direction. Therefore, in the region where the maximum influence degree is greater than 50%, the C color is selected according to the influence degree, and in the region where the influence degree is 50% or less, the A and B colors which are the nearest pixels regardless of the influence degree. By selecting, an image maintaining orthogonality can be generated.

  Next, a process for maintaining linearity and orthogonality in orthogonal portions where four background colors are adjacent to each other will be described. FIG. 40 is an explanatory diagram showing coordinates of each pixel and interpolation pixel of an input image having four color orthogonal portions, and FIG. 41 is an explanatory diagram showing an example of the same attribute layer information in the case of having four color orthogonal portions. FIG. 42 is an explanatory diagram illustrating an example of the degree of influence in the case of having four color orthogonal portions, and FIG. 43 is an explanatory diagram illustrating an example of an interpolated pixel by high resolution processing in the case of having four color orthogonal portions. It is. In FIG. 40, white circles indicate the coordinates of each pixel of the input image data, and black dots indicate the coordinates of the interpolation pixel after the resolution enhancement processing that is the output image data, and indicate the relationship between the positions. A, B, C, and D in white circles indicate the color (feature amount) of each pixel. In the example of FIG. 40, a state where the regions represented by colors A, B, C, and D are separated by a straight line in the vertical direction and a straight line in the horizontal direction is shown. In addition, each pixel of the center 2 × 2 pixel lattice is the target pixel (1, 1), (1, 2), (2, 1), (2, 2).

  As shown in FIGS. 40 to 43, in the portion where the four colors are orthogonal, there are four pixels for each of the A, B, C, and D colors, and there is no interpolation target coordinate that has a maximum influence level of 50%. For this reason, in all the interpolation target coordinates, the nearest pixel is selected and interpolated to generate an image maintaining orthogonality.

  Next, a method for extracting thin line layer information will be described. A pixel of interest belonging to an oblique thin line with a width of 1 pixel or an isolated point has a very small influence value compared to the influence degree of the pixels of the background image. There are times when thinning or thin lines disappear. In order to prevent these, first, it is determined whether or not the pixel of interest belongs to a thin line such as an oblique thin line or an isolated point.

  44 and 45 are explanatory diagrams showing an example of the fine line layer information. The fine line layer information is information indicating whether or not the target pixel belongs to the fine line. When obtaining the fine line layer information illustrated in FIGS. 44 and 45, it is assumed that the input image data is as shown in FIG. That is, it is assumed that 4 × 4 pixel input image data (low-resolution raster data) includes pixels of three colors, an A color pixel, a B color pixel, and a C color pixel. The feature amount of each pixel corresponds to, for example, a density value for a monochrome (monochrome) image, a density value for each color element such as RGB or CMYK for a color image, and further, an attribute (character, vector graphics) of each pixel. Additional information such as tag information indicating a photo area) can also be added. In this embodiment, it is assumed that pixels having the same density value for all color elements have the same feature amount (the pixel values are the same).

  44 (a), 44 (b), 45 (c), and 45 (d) are respectively four pixels of interest (1, 1), (1) of the 2 × 2 pixel grid shown in the example of FIG. 2), (2, 1), and (2, 2), a method for obtaining four sets of fine line layer information corresponding to each. In FIG. 19 and FIG. 20, a numerical value indicated in a circle indicating each pixel indicates a distance from the pixel of interest, and the identity of the feature amount and the continuity of the pixels in order from the pixel group having the smaller numerical value (distance). Perform sex determination. The determination of continuity is performed when the pixel of interest or the already-determined pixel of the same attribute exists (adjacent) in the upper, lower, left, and right four pixels (excluding vertical and horizontal directions and diagonal directions) of the determination target pixel. Determined to be continuous. By repeating these processes from the pixel group at the distance 1 to the pixel group at the distance 5, the determination is made for all the pixels. In FIG. 44 and FIG. 45, pixels that are determined to have the same attribute as a result are indicated by black circles, and pixels that are determined not to have the same attribute are indicated by white circles. Further, the difference from the determination of the presence or absence of the same attribute in FIGS. 4 and 5 is that only four pixels on the top, bottom, left, and right are used as reference pixels when determining continuity.

  Next, a procedure for determining the presence / absence of the same attribute for obtaining the thin line layer information will be described. For the pixel group whose distance from the target pixel (1, 1) is 1, as Step 1, the pixel (0, 1) is different from the target pixel (1, 1) and is continuous (adjacent). Therefore, it is determined that the attribute is not the same (no attribute is the same). Hereinafter, similarly, since the pixel (1, 0) is different in color from the target pixel (1, 1) and is continuous, it is determined to have the same attribute. Since the pixel (1, 2) is different from the target pixel (1, 1) and is continuous, the pixel (1, 2) is determined to have the same attribute. Since the pixel (2, 1) is different from the target pixel (1, 1) and is continuous, the pixel (2, 1) is determined to have the same attribute.

  Next, for the pixel group whose distance from the target pixel (1, 1) is 2, as Step 2, the pixel (0, 0) has the same color as the target pixel (1, 1), and the continuous pixels are Because there is no attribute, it is determined that the attribute is not the same. The pixel (0, 2) has a different color from the target pixel (1, 1) and has no continuous pixel, and thus is determined to have the same attribute. The pixel (2, 0) is different in color from the target pixel (1, 1) and has no continuous pixel, and thus is determined to have the same attribute. The pixel (2, 2) has the same color as the pixel of interest (1, 1) and has no continuous pixel, so it is determined to have the same attribute.

  Next, with respect to the pixel group whose distance from the target pixel (1, 1) is 3, as Step 3, the pixel (1, 3) is different in color from the target pixel (1, 1), and the continuous pixels are Because there is no attribute, it is determined that the attribute is not the same. The pixel (3, 1) is different in color from the target pixel (1, 1) and has no continuous pixel, so it is determined to have the same attribute.

  Next, with respect to the pixel group whose distance from the target pixel (1, 1) is 4, as Step 4, the pixel (0, 3) is different in color from the target pixel (1, 1), and the continuous pixels are Because there is no attribute, it is determined that the attribute is not the same. The pixel (2, 3) is different in color from the target pixel (1, 1) and has no continuous pixel, and therefore is determined to have the same attribute. The pixel (3, 0) is different in color from the target pixel (1, 1) and has no continuous pixel, so it is determined to have the same attribute. The pixel (3, 2) has a different color from the target pixel (1, 1) and has no continuous pixel, and thus is determined to have the same attribute.

  Next, with respect to the pixel group whose distance from the target pixel (1, 1) is 5, as Step 5, the pixel (3, 3) is the same color as the target pixel (1, 1), and the continuous pixel is Because there is no attribute, it is determined that the attribute is not the same.

  This completes the determination of the same attribute for all pixels. Based on this determination result, the number of pixels obtained by adding the target pixel itself to the number of pixels determined to have the same attribute (with the same attribute) is counted, and when the counted number of the same attribute pixel is 1 (only the target pixel), the target pixel Are isolated and belong to a fine line (thin line layer). If the number of pixels with the same attribute is 2 or more, it is determined that the target pixel does not belong to the fine line. By performing the above-mentioned processing similarly for the target pixels (1, 2), (2, 1), (2, 2), it is possible to obtain four thin line layer information corresponding to the four target pixels.

  44 and 45, the pixel of interest (1, 1), (2, 2) is determined to belong to a thin line because both of the same attribute pixels are 1, and the pixels of interest (1, 2), ( 2 and 1) are determined not to belong to the thin line because the same attribute pixel is 2 or more.

  Next, a process in which a thin line layer information extraction process is added to the basic process of the resolution enhancement process shown in FIGS. 7 to 10 will be described. 46 is an explanatory diagram showing coordinates of each pixel and interpolation pixel of the input image, FIG. 47 is an explanatory diagram showing an example of the same attribute layer information, FIG. 48 is an explanatory diagram showing an example of the degree of influence, FIG. 49 is an explanatory diagram showing an example of the fine line layer information, and FIG. 50 is an explanatory diagram showing an example of the interpolated pixel by the high resolution processing. In FIG. 46, white circles indicate the coordinates of each pixel of the input image data, and black dots indicate the coordinates of the interpolation pixel after the high resolution processing that is the output image data, and indicate the relationship between the positions. A and B in white circles indicate pixel colors (features), respectively. In addition, each pixel of the center 2 × 2 pixel lattice is the target pixel (1, 1), (1, 2), (2, 1), (2, 2).

  As shown in FIG. 49, regarding the target pixel (1, 1), (1, 2), (2, 1), the number of pixels with the same attribute (count) is 14, which is 2 or more. However, for the target pixel (2, 2), the number of pixels with the same attribute (count) is 1, indicating that the target pixel (2, 2) belongs to a thin line (thin line layer).

  Based on this information, when the influence level of the target pixel (2, 2) is equal to or greater than the second threshold, the target pixel (2, 2) determined to belong to the thin line is preferentially selected as the copy source pixel. In the influence degree by the pixel of interest (2, 2) shown in FIG. 48D, if the second threshold is 24, for example, the influence at the lower right interpolation target coordinate among the interpolation target coordinates of 4 × 4 pixels. Since the degree is 25 and larger than the threshold (22), the pixel value (B) of the target pixel (2, 2) is copied as the interpolation pixel of the lower right interpolation target coordinate. FIG. 50C is an example in which a pixel having an influence degree of the pixel of interest (2, 2) having a second threshold value or more is selected, and a pixel in which B is indicated in a circle is a priority selection having the second threshold value or more. Pixels that are indicated by “x” in a circle are pixels that have an influence degree less than the second threshold and are normally selected pixels. For this normally selected pixel portion, the pixel selected by the basic processing shown in FIG. 50A is applied, and finally the final selected pixel shown in FIG. 50B is obtained.

  Note that the absolute value of the degree of influence of the pixel of interest in the fine line layer varies greatly depending on the length (number of pixels) of the fine line. Therefore, when the threshold is set to a fixed value, the line width increases when the fine line is long, and the fine line is short. Sometimes an output image with a reduced line width is obtained. In order to keep the line width of the output image as constant as possible, it is desirable to adjust the threshold according to the length of the thin line.

  Specifically, from the threshold table of the fine line influence level in which a plurality of second threshold values prepared in advance are set, an appropriate first value is determined depending on the count value of the same attribute layer of the target pixel of interest (that is, the number of pixels having the same attribute). 2 Select a threshold.

  FIG. 51 shows an example of a threshold table of the fine line influence degree. Since the threshold value for reproducing an appropriate line width varies depending on the interpolation magnification and the characteristics of the image forming apparatus (for example, a printer), for example, it is necessary to experimentally determine and set an appropriate value. In the example of FIG. 51, the case where the interpolation magnification is 2 × 2 and 4 × 4 are illustrated, but the interpolation magnification is not limited to this. As the count number of the same attribute layer, that is, the thin line length increases, the influence of the difference of one count on the width of the thin line tends to decrease. Therefore, in this embodiment, when the count number is 3 or more The same threshold is used.

  In the example of FIG. 47D, since the interpolation magnification is 4 × 4 and the count value of the same attribute layer of the target pixel (2, 2) is 2, 24 is selected as the second threshold value.

  If the thin line layer processing as described above is not performed, an interpolation pixel as shown in FIG. 50A is obtained, and one target pixel (2 in the 2 × 2 pixel grid of the input image shown in FIG. 46) is obtained. It can be seen that the feature 2) (pixel value B) has disappeared. By performing the thin line layer processing, it is possible to suppress the thin line from becoming thinner and the thin line from disappearing.

  Next, a description will be given of a case where tag information indicating a pixel region attribute is used in addition to color information when determining the feature amount of each pixel. As tag information, for example, various attribute information such as basic attribute information such as a photo area, a character area, a vector graphics area, or a more detailed classification thereof, or a binary or halftone classification is used. Can do. In the present embodiment, two attributes of a character area and a vector graphics area will be described as an example.

  FIG. 52 is an explanatory diagram showing the coordinates of each pixel and interpolation pixel of a two-color input image composed of a character area and a vector graphics area. In FIG. 52, white circles indicate the coordinates of each pixel of the input image data, and black dots indicate the coordinates of the interpolation pixel after the resolution enhancement processing that is the output image data, and indicate the relationship between the positions. A in a white circle indicates a pixel in the character area in A color, B indicates a pixel in the character area in B (b) color, and b indicates a pixel in the vector graphics area in B (b) color. In addition, each pixel of the central 2 × 2 pixel grid is the target pixel (1, 1), (1, 2), (2, 1), (2, 2).

  FIG. 53 is an explanatory diagram showing an example of the same attribute layer information when tag information is not used, in which the same attribute layer information is extracted using only color information. FIG. 54 is an explanatory diagram showing an example of an interpolated pixel obtained by high resolution processing when tag information is not used. As shown in FIG. 54, each pixel B and b of the input image is treated as having the same attribute.

  FIG. 55 is an explanatory diagram showing an example of the same attribute layer information when tag information is used. The attribute layer information is extracted assuming that the character area and the vector graphics area are different attributes. . FIG. 56 is an explanatory diagram showing an example of an interpolation pixel by resolution enhancement processing when tag information is not used. As shown in FIG. 56, the B and b pixels are treated as having the same attribute.

  That is, as shown in FIG. 54, when tag information is not considered, the pixels of B and b are treated as having the same attribute, and the corner portion of the A color region is rounded. On the other hand, as shown in FIG. 56, when tag information is taken into consideration, it is possible to obtain an image in which orthogonality is ensured as it should be.

  FIG. 57 is an explanatory diagram showing a part of an image after the high resolution processing of the present invention. FIG. 57 shows the original image shown in FIG. 60 which has been subjected to the high resolution processing of the present invention. As shown in FIG. 57, according to the present invention, compared to the conventional example (for example, FIG. 61, FIG. 62, FIG. 63, etc.), there is no occurrence of density or color blur in the contour portion, and jaggies are reduced. A reduced high resolution image can be obtained.

  58 and 59 are flowcharts showing the processing procedure of the high resolution processing of the present invention. The high resolution processing unit 212 (hereinafter referred to as “processing unit 212”) acquires low resolution raster data, sets initial interpolation target coordinates (S11), and 2 × around the pixel of the set interpolation target coordinates. Two pixels are selected as the target pixel (S12).

  The processing unit 212 extracts surrounding 4 × 4 pixels (reference pixels) including the target pixel as input data (S13), and refers to the reference pixel to obtain the same attribute layer information (continuous in eight directions) for one target pixel. Calculate (S14). The processing unit 212 counts the number of pixels in the same attribute layer (S15), and performs interpolation processing on the calculated same attribute layer information to calculate interpolation layer information (S16).

  The processing unit 212 calculates the degree of influence of the pixel of interest at the interpolation target coordinates by performing a predetermined filter process on the interpolation layer information (S17). The processing unit 212 calculates a fine line layer (continuous in four directions) (S18), and counts the number of pixels in the fine line layer (S19). The calculation of the thin line layer is not necessarily an essential process, but by performing this process, a high-resolution image with even better image quality can be obtained as described above.

  The processing unit 212 determines whether or not all the processing of the 2 × 2 target pixel has been completed (S20). If not all processing has been completed, the processing from step S14 is performed, and the same processing is performed for each target pixel. Repeat the process. When all the processes are completed (YES in S20), the processing unit 212 determines whether there is a target pixel having a fine line layer count value of 1 (S21).

  When there is no target pixel having a thin line layer count value of 1 (NO in S21), the processing unit 212 selects a target pixel having the maximum influence (S22), and whether or not the maximum influence is less than 50%. If it is not less than 50% (NO in S23), the target pixel having the maximum influence is selected as a copy source pixel (S24).

  When there is a target pixel whose thin line layer count value is 1 (YES in S21), the processing unit 212 selects a threshold value based on the same attribute layer count value of the target pixel (S25), and the influence degree of the target pixel is the threshold value It is determined whether or not this is the case (S26). If the degree of influence is not equal to or greater than the threshold (NO in S26), the processing unit 212 performs the process of step S22. If the degree of influence is equal to or greater than the threshold (YES in S26), the target pixel is used as a copy (copy) source pixel. Select (S27). When the maximum influence degree is less than 50% (YES in S23), the processing unit 212 selects the target pixel closest to the interpolation target coordinate as a copy source pixel (S28).

  The processing unit 212 copies the feature amount of the selected target pixel to the interpolation target coordinates (S29), and determines whether all the processes of the M × N interpolation target coordinates are finished (S30). If all the processing of the interpolation target coordinates has not been completed (NO in S30), the processing unit 212 continues the processing from step S21.

  When all the processes for the interpolation target coordinates are completed (YES in S30), the processing unit 212 determines whether all the processes for the image area (entire input image) are completed (S31), and the image area (input) Entire image) If all processes have not been completed (NO in S31), the process from step S11 is continued, and if all processes in the image area (entire input image) have been completed (YES in S31), the process is terminated. .

  As described above, according to the present invention, the pixel value of the interpolation pixel is calculated based on the pixel value (feature value) of the target pixel group having the same influence degree around the interpolation target coordinates. Density and color blur do not occur, and an interpolation result suitable for characters and text graphics can be obtained. In addition, since a set of target pixel groups is selected according to the degree of influence of each target pixel group around the interpolation target coordinates, smoothing processing can be performed on the contour shape of the drawing element, and the degree of influence can be calculated. Since the information to be used (for example, pixel values) is information of the input image (original image) before the interpolation process, a good smoothing result can be obtained regardless of the size of the high resolution ratio. Further, it is possible to obtain a high-resolution image free from jaggy and free from density blur and color mixing. Further, since the boundary of the drawing element is detected based on the difference in pixel value (feature amount) between the target pixels, pixel attribute information such as TAG information is not required for contour detection.

  In addition, according to the present invention, when an input image in which an image composed of pixels of the same color such as a character having a diagonal shape or a curved edge shape is arranged on a gradation background is subjected to high resolution conversion, the diagonal edge portion Or generation | occurrence | production of the jaggy of a curve edge part is suppressed and the favorable holding | maintenance of an edge shape is attained.

  Further, according to the present invention, in the smoothing process using resolution conversion, when performing the same attribute determination of the pixel of interest and the surrounding pixel group of the pixel of interest, and determining the pixel value of the interpolation pixel based on the determination result, By calculating interpolated pixel values from the pixel values of a plurality of target pixels belonging to the selected same attribute layer, for images that contain natural images such as characters and artificial images such as characters and vector graphics, A good image can be obtained by performing both the smoothing process for the contour shape of the artificial image and the smoothing process for the density of the natural image.

  Further, according to the present invention, when calculating the degree of influence at the coordinates to be interpolated for each target pixel, the number of pixels to be referred to (the number of reference pixels) can be increased, and a wider range of information is taken into consideration. Thus, an image with a smoother contour can be obtained. Further, when determining the presence or absence of the same attribute of the pixel of interest and its surrounding reference pixels, if the feature amount differs by determining whether the feature amount of the pixel partially matches or completely matches, the attribute is Even when similar color image areas are adjacent to each other, each image area can be correctly recognized as a different area, and smoothing can be performed on each outline.

  Further, according to the present invention, the reference pixel has the same attribute according to whether or not the reference pixel is continuous with the target pixel through the target pixel or a pixel that has already been determined to have the same attribute in eight vertical and horizontal diagonal directions. Recognizing two areas of the same color as separate areas even if there is a thin line between two areas of the same color, such as a thin line drawn on the background color, by determining the presence or absence Thus, the absolute value of the degree of influence of each region can be reduced, and the thin line can be prevented from becoming extremely thin or the thin line from disappearing.

  Further, according to the present invention, when selecting a pixel of interest corresponding to the interpolation target coordinates, by selecting the pixel of interest having the greatest degree of influence among the pixels of interest, there is no density or color blur in the contour portion, Interpolation results suitable for characters and text graphics can be obtained. In addition, smoothing processing can be performed on the contour shape of the drawing element, and a high-resolution image free from jaggy can be obtained.

  Further, according to the present invention, when the degree of influence is compared with the first threshold value and the degree of influence is smaller than the first threshold value, the interpolation target coordinates having a small degree of influence are selected by selecting the target pixel closest to the interpolation target coordinates. In contrast, by suppressing the selection of the pixel of interest having the greatest degree of influence and selecting the pixel of interest closest to the coordinates to be interpolated as the pixel of interest of the copy source, the intersection of the vertical and horizontal straight lines For example, by detecting a portion where drawing elements of three or more colors are orthogonal to each other and avoiding the smoothing process, linearity and orthogonality at the orthogonal portion can be maintained.

  In addition, according to the present invention, it is determined whether or not the target pixel belongs to a diagonal line having a width of one pixel or a thin line that is an isolated point. If the degree of influence at the target coordinates is larger than the second threshold value by comparing with the two threshold values, the target pixel determined to belong to the thin line is selected in association with the interpolation target coordinates. If the degree of influence at the interpolation target coordinates is smaller than the second threshold, the pixel of interest having the highest degree of influence is selected in association with the interpolation target coordinates. Thereby, it is possible to prevent the oblique thin lines and the isolated points from disappearing by extracting the oblique thin lines and the isolated points and preferentially considering the degree of influence thereof.

  Further, according to the present invention, by changing the second threshold according to the number of pixels having the same attribute as the target pixel determined to belong to the thin line, the line width after processing is made constant regardless of the length of the thin line. Can keep.

  In the above-described embodiment, it is determined whether or not the target pixel belongs to a diagonal line of one pixel width or a thin line that is an isolated point, and the degree of influence on the interpolation target coordinates of the target pixel determined to belong to the thin line and the second When the influence level at the interpolation target coordinates is smaller than the second threshold value by comparing with the threshold value, the pixel of interest having the greatest influence level is selected in association with the interpolation target coordinates. It may be. That is, when the influence degree in the interpolation target coordinates is smaller than the second threshold value, the influence degree of the four target pixels in each interpolation target coordinate is compared with the first threshold value, and the interpolation target coordinates having a maximum influence degree of less than 50%. For this, it is also possible to select the pixel of interest closest to the coordinates to be interpolated as the pixel of interest of the copy source, and select the pixel of interest having the greatest influence when the influence is greater than or equal to the first threshold. it can. By performing such a process, it is possible to perform a smoothing process on the contour shape while maintaining the reproducibility of oblique thin lines and isolated points and the linearity and orthogonality at the orthogonal part.

  In the above-described embodiment, the image forming apparatus is not limited to a single printer, but may be a digital multi-function peripheral having a copy function and a fax function in addition to a printer function.

  The present invention can also be realized as a configuration in which the above-described image processing procedure for increasing the resolution is recorded on a computer-readable recording medium in which a program to be executed by a computer is recorded. As a result, a recording medium on which program codes (execution format program, intermediate code program, source program) for performing high resolution processing are recorded can be provided in a portable manner.

  In the present embodiment, the recording medium may be a memory (not shown) such as a ROM itself, which is processed by a microcomputer, or may be a program medium. As the external storage device, a program reading device may be provided, and a program medium that can be read by inserting a recording medium therein may be used.

  In any case, the stored program may be configured to be accessed and executed by the microprocessor, or in any case, the program code is read and the read program code is displayed on the microcomputer. The program may be downloaded to an unprogrammed program storage area and executed. It is assumed that this download program is stored in the main device in advance.

  Here, the program medium is a recording medium configured to be separable from the main body, such as a tape system such as a magnetic tape or a cassette tape, a magnetic disk such as a flexible disk or a hard disk, a CD-ROM, MO, MD, DVD, or the like. Semiconductors such as optical discs, IC cards (including memory cards) / optical cards, etc., or mask ROM, EPROM (Erasable Programmable Read Only Memory), EEPROM (Electrically Erasable Programmable Read Only Memory), flash ROM, etc. It may be a medium that carries a fixed program code including a memory.

  In the present embodiment, since the system configuration is capable of connecting to a communication network including the Internet, the medium may be a medium that dynamically carries the program code so as to download the program code from the communication network. When the program is downloaded from the communication network in this way, the download program may be stored in the main device in advance or installed from another recording medium. The present invention can also be realized in the form of a computer data signal embedded in a carrier wave in which the program code is embodied by electronic transmission. The image processing method described above is executed by reading the recording medium by a program reading device provided in a computer system.

2 Printer 21 Image Processing Unit 211 Raster Data Generation Unit 212 High Resolution Processing Unit 2121 Same Attribute Layer Extraction Unit 2122 Layer Information Interpolation Unit 2123 Layer Information Filter Unit 2124 Fine Line Layer Extraction Unit 2125 Copy Source Pixel Selection Unit 2126 Pixel Feature Amount Copying Unit 213 Halftone processing unit 22 Printing engine unit

Claims (16)

In an image processing apparatus that generates an output image by performing an interpolation process on an input image,
An influence degree calculating means for calculating an influence degree in the interpolation target coordinates of each of a plurality of target pixels around the plurality of interpolation target coordinates;
Selection means for selecting a set of target pixel groups having the same influence level from the target pixels in association with each interpolation target coordinate according to the influence level calculated by the influence level calculation means;
An image processing apparatus comprising: an interpolation pixel value calculating unit that calculates a pixel value of an interpolation pixel at an interpolation target coordinate based on a pixel value of a target pixel group selected by the selection unit. For each target pixel, the same attribute determination unit that determines the presence or absence of the same attribute of each pixel in the pixel block including the target pixel and the plurality of target pixels;
Extraction means for extracting the same attribute information assigned in association with each pixel based on the determination result determined by the same attribute determination means;
Same attribute information interpolation means for interpolating the same attribute information extracted by the extraction means at a predetermined interpolation magnification,
The influence calculating means is
The configuration is such that, for each pixel of interest, the degree of influence by surrounding pixels existing around the interpolation target coordinates is calculated using the same interpolation attribute information interpolated by the same attribute information interpolation means. Item 8. The image processing apparatus according to Item 1. The attribute determination means includes
3. The configuration according to claim 2, wherein for each target pixel, the presence or absence of the same attribute is determined according to whether or not the pixel values of the target pixel and each pixel in the pixel block match. An image processing apparatus according to 1. The attribute determination means includes
The presence or absence of the same attribute is determined for each target pixel depending on whether or not the pixel value of each pixel in the pixel block is included in a pixel value range provided for the pixel value of the target pixel. The image processing apparatus according to claim 2, wherein the image processing apparatus is configured. The attribute determination means includes
For each pixel of interest, the pixel value range is selected according to the image attribute value of the pixel of interest, and depending on whether the pixel value of the pixel of interest is included in the pixel value range and whether the image attribute value is the same Is configured to determine the presence or absence of the same attribute,
The interpolation pixel value calculation means includes
The pixel value of the interpolated pixel is calculated from a pixel value of the target pixel group selected by the selection unit according to a procedure selected according to an image attribute value of the target pixel group. Item 5. The image processing apparatus according to Item 4. The attribute determination means includes
For each pixel of interest, whether each pixel in the pixel block is adjacent to the pixel of interest through the pixel of interest or a pixel that has already been determined to have the same attribute is determined. The image processing apparatus according to claim 3, wherein the image processing apparatus is configured. The selection means includes
7. The configuration according to claim 2, wherein the pixel of interest having the greatest influence degree calculated by the influence degree calculating unit is selected for each interpolation target coordinate. 8. The image processing apparatus described. A first comparing means for comparing the influence degree calculated by the influence degree calculating means with a first threshold;
The selection means includes
8. The configuration according to claim 2, wherein when the degree of influence at the interpolation target coordinates is smaller than the first threshold value, the pixel of interest closest to the interpolation target coordinates is selected. The image processing apparatus according to claim 1. Fine line determination means for determining whether or not the pixel of interest belongs to a diagonal line or a thin line that is an isolated point;
A second comparing means for comparing the degree of influence of the target pixel determined to belong to the thin line by the thin line determining means at the interpolation target coordinates and a second threshold value;
The selection means includes
When the degree of influence at the interpolation target coordinates is larger than the second threshold, the pixel of interest determined to belong to the thin line is selected in association with the interpolation target coordinates,
The configuration is such that, when the degree of influence at the interpolation target coordinates is smaller than the second threshold value, the pixel of interest having the highest degree of influence is selected in association with the interpolation target coordinates. The image processing apparatus according to claim 1. Fine line determination means for determining whether or not the pixel of interest belongs to a diagonal line or a thin line that is an isolated point;
A second comparing means for comparing the degree of influence of the target pixel determined to belong to the thin line by the thin line determining means at the interpolation target coordinates and a second threshold value;
The selection means includes
When the degree of influence at the interpolation target coordinates is larger than the second threshold, the pixel of interest determined to belong to the thin line is selected in association with the interpolation target coordinates,
The first comparing means includes
When the influence degree at the interpolation target coordinates is smaller than the second threshold value, the influence degree calculated by the influence degree calculation means is configured to be compared with the first threshold value,
When the degree of influence at the interpolation target coordinates is smaller than the first threshold, the pixel of interest closest to the interpolation target coordinates is selected,
9. The configuration according to claim 8, wherein when the degree of influence at the interpolation target coordinates is equal to or greater than the first threshold, the pixel of interest having the largest degree of influence calculated by the influence degree calculating means is selected. The image processing apparatus described. The thin line determination means includes
When the same attribute determining unit determines that the target pixel does not have the same attribute as the pixel adjacent to the target pixel in the vertical or horizontal direction, it is configured to determine that the target pixel belongs to a thin line. The image processing apparatus according to claim 9 or 10.   The change means for changing the second threshold value according to the number of pixels having the same attribute as the target pixel determined to belong to the fine line by the fine line determination means. The image processing apparatus according to item 1.   An image forming apparatus comprising: an image forming unit that forms an output image; and the image processing apparatus according to any one of claims 1 to 12. In a computer program for causing a computer to function as means for generating an output image by performing interpolation processing on an input image,
Computer
An influence degree calculating means for calculating an influence degree in the interpolation target coordinates of each of a plurality of target pixels around the plurality of interpolation target coordinates;
In accordance with the calculated influence degree, a selection unit that selects a set of attention pixel groups having the same influence degree from the attention pixels in association with each interpolation target coordinate;
A computer program that functions as an interpolation pixel value calculation unit that calculates a pixel value of an interpolation pixel at an interpolation target coordinate based on a pixel value of a target pixel group selected by the selection unit.   15. A computer-readable recording medium on which the computer program according to claim 14 is recorded. In an image processing method by an image processing apparatus that performs an interpolation process on an input image to generate an output image,
Calculating the degree of influence in the interpolation target coordinates of each of the plurality of target pixels around the plurality of interpolation target coordinates;
In accordance with the calculated influence level, a set of target pixel groups having the same influence level is selected from the target pixels in association with each interpolation target coordinate,
An image processing method, comprising: calculating a pixel value of an interpolation pixel at an interpolation target coordinate based on a pixel value of a selected pixel group of interest.

Images