JP2004289340A - Image enlargement processing device - Google Patents

Abstract

An object of the present invention is to smooth a high-quality image at any magnification.
In an image processing apparatus having a scaling function for enlarging image data, a line FIFO for storing one line of image data, a storage memory for each post-enlargement pixel corresponding to a maximum enlargement ratio, and a memory. A memory 103 for holding pixel data before and after the pixel data of interest, a counter 104 for counting the number of pixels according to the enlargement ratio, and an inter-pixel interpolation data processing unit 106 for performing image smoothing processing according to the enlargement ratio. In this configuration, the smoothing process according to the enlargement ratio is performed instead of the uniform smoothing process for several pixels only near the boundary.
[Selection diagram] Fig. 1

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a technology for enlarging digital image data of a digital copying machine, a printer, a digital camera, and the like, and more particularly to an image enlarging process technology suitable for performing a high-quality image smoothing process on an enlarged image. is there.
[0002]
[Prior art]
In a conventional image enlargement processing apparatus, image data read by a scanner in units of one line is temporarily stored in a line FIFO (First-In First-Out, Fifo), and thereafter, from the FIFO in accordance with the enlargement ratio. The required enlarged image is generated by controlling the reading of the image data.
[0003]
For example, in the case of enlargement by 2 times in the main scanning and 2 times in the sub-scanning, the data of the first pixel read from the FIFO is first repeated twice in the main scanning direction to generate twice the image data in the horizontal direction, thereby producing one line. The data is sequentially stored in the output FIFO of the eye.
[0004]
In the next line, the same image data is read from the same line FIFO and is repeated twice in the main scanning direction to generate twice the image data in the horizontal direction, and sequentially stores the enlarged image data of the second line in the output FIFO.
[0005]
As a result, an image which is twice as large in the main scanning and the sub-scanning is generated by this processing. By repeating this processing for the data of the second and subsequent pixels, an enlarged image for all the image data is generated.
[0006]
In this conventional technique, there is no correlation between the enlarged image data generated for each original image data, and depending on the image data, there may be a sharp change when considered as a frequency component at the boundary. As a result, a harmonic component becomes conspicuous.
[0007]
As a conventional technique for coping with such a problem, there is a technique for performing a smoothing process for providing smoothness of an image near a boundary by uniformly performing a convolution operation on several pixel data at the boundary.
[0008]
However, in this smoothing processing technique, a portion where a harmonic component is conspicuous still occurs in an image other than the vicinity of the boundary. In particular, the symptoms become more pronounced as the magnification increases.
[0009]
Conversely, in the case where it is necessary to emphasize the edges to make the character image clear like a character, the character may be blurred if the smoothing process is performed uniformly as described above.
[0010]
As a technique for reducing the blur of an image by clarifying the boundary of an edge portion of an image, for example, as described in Patent Document 1, interpolation is performed by giving a nonlinear characteristic to a coefficient by which original image data is multiplied. There is a technique for enlarging an image signal.
[0011]
Further, as a technique for improving the smoothness of the image near the boundary and reducing the blur of the image when the image is enlarged in the character / line drawing portion, there are techniques described in Patent Documents 2 and 3, for example.
[0012]
According to the technique described in Patent Document 2, when an image is scaled and output, appropriate scaling processing is performed on each of the character / line image portion and the pseudo halftone component, and the resultant is synthesized. Further, in Patent Document 3, while the technology described in Patent Document 2 is for binary images, an edge portion and a non-edge portion of a multi-valued image are enlarged by an enlargement method suitable for both, A technique for performing synthesis to create a high-quality enlarged image is described.
[0013]
However, these conventional techniques cannot appropriately perform smoothing processing on an image other than near the boundary.
[0014]
[Patent Document 1]
JP-A-2000-36041
[Patent Document 2]
JP-A-7-121692
[Patent Document 3]
JP 2000-99713 A
[0015]
[Problems to be solved by the invention]
The problem to be solved is that the conventional technology cannot efficiently perform smoothing processing on an image other than the vicinity of the boundary, and in particular, the quality of the enlarged image deteriorates as the enlargement ratio increases. Is a point.
[0016]
SUMMARY OF THE INVENTION An object of the present invention is to solve these problems of the prior art and to enable output of a high-quality smooth enlarged image at any enlargement ratio.
[0017]
[Means for Solving the Problems]
In order to achieve the above object, the present invention provides an image enlargement processing device that enlarges image data, by performing not a uniform smoothing process for a few pixels near the boundary but a smoothing process in accordance with an enlargement ratio. The enlargement is performed as an image having a gradation closer to the original image. For example, as shown in FIG. 1 described later, each interpolation for interpolating between a target pixel and an adjacent pixel for enlargement is performed. An inter-pixel interpolation data processing unit (106) for calculating the value of a pixel according to the value of the pixel of interest and the adjacent pixel and the enlargement ratio is provided, and an image smoothing process according to the enlargement ratio is performed. As the configuration of the inter-pixel interpolation data processing unit (106), for example, as in an inter-pixel interpolation data processing unit (206) shown in FIG. And a divider (206c) that divides the difference value obtained by the difference calculator (206b) by an enlargement ratio, and a divider (206b) that calculates the difference between the value of the pixel of interest and the value of the pixel after the pixel of interest. An arithmetic unit (multiplier 206d) for changing the value of the division result of (206c) according to the position of each interpolation pixel, and using the value changed by this arithmetic unit (206d), the value of each interpolation pixel at that position is calculated. A calculator (206e) for calculating, or zero data insertion for inserting zero data into each interpolated pixel between a target pixel and an adjacent pixel as in an inter-pixel interpolation data processing unit (306) shown in FIG. Container (306a Oversampling filter by the product-sum operation of the value of the pixel of interest and the value of the adjacent pixel, and each zero data between the pixel of interest and the adjacent pixel, and a predetermined coefficient (306b) corresponding to the number of interpolation pixels An oversampling filter (306c) for performing processing is provided. Furthermore, by determining the presence or absence of smooth processing according to the type of image, smooth processing is not performed on an image such as a character image in which edges must be emphasized, and enlargement is performed as a clearer image. As a possible configuration, as shown in FIG. 4, a determination unit (414) that determines whether image data to be enlarged is a character image, and a determination that the determination unit (414) is not a character image If the result is a result, a selector (415) for selecting and executing the image smoothing process by the inter-pixel interpolation data processing unit (406) is provided. An edge determination unit (412) that performs an edge determination and an isolated point determination unit (413) that performs an isolated point determination on the image data to be enlarged are processed based on the determination results. There is a determining configuration whether the character image.
[0018]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
[0019]
FIG. 1 is a block diagram showing a first configuration example of an image enlargement processing device according to the present invention, FIG. 2 is a block diagram showing a second configuration example of the image enlargement processing device according to the present invention, and FIG. FIG. 4 is a block diagram showing a third configuration example of the image enlargement processing device according to the present invention, and FIG. 4 is a block diagram showing a fourth configuration example of the image enlargement processing device according to the present invention.
[0020]
In FIG. 1, 101 is a line FIFO for storing one line of image data, 102 is a post-enlargement storage data storage memory for storing pixels after each enlargement process corresponding to the maximum enlargement ratio, and 103 is target pixel data. , A pixel counter 104 for counting the number of pixels for specifying a memory area according to an enlargement ratio (described as “pixel counter for specifying a memory area” in the figure), 105 Reference numeral denotes an enlargement ratio setting register for holding an enlargement ratio designated by a CPU (Central Processing Unit). Reference numeral 106 denotes an inter-pixel interpolation data processing unit (in FIG. Interpolation data processing block ") and 107 is an output line F for outputting image data after the enlargement processing. Is a FO.
[0021]
As described above, the image enlargement processing device of the example shown in FIG. 1 includes the inter-pixel interpolation data processing unit 106, and when the image data is enlarged, the inter-pixel interpolation data processing unit 106 By calculating the value of each interpolated pixel to be interpolated between a pixel and an adjacent pixel according to the value of the pixel of interest and the adjacent pixel and the enlargement ratio, the image smoothing process according to the enlargement ratio is efficiently performed. be able to. Hereinafter, the operation will be described.
[0022]
That is, the image enlargement processing apparatus of this example first sets the enlargement ratio in each of the main scanning direction and the sub-scanning direction in each setting register 105. Then, one line of image data read from a scanner or the like is loaded into the line FIFO 101.
[0023]
When image data is read from the line FIFO 101, a total of two pixels, that is, the pixel of interest to be enlarged and the pixel behind it, are read and stored in the pixel storage memory 103. These image data are used when calculating an interpolation pixel value later.
[0024]
The pixel count 104 calculates the number of continuous pixels after the enlargement according to the enlargement ratio set in the setting register 105, and stores the data after the enlargement processing according to the number in the post-enlargement storage data storage memory 102. Determine the memory area.
[0025]
The target pixel data stored in the pixel storage memory 103 is stored at the start address of the area in the post-enlargement storage data storage memory 102 determined in this way.
[0026]
Then, according to the counting result of the pixel counter 104 corresponding to the enlargement ratio, the pixel data interpolated by the inter-pixel interpolation data processing unit 106 is sequentially read from the next address of the target image data in the post-enlargement storage data storage memory 102. Then, save.
[0027]
Note that the inter-pixel interpolation data processing unit 106 of this example is not a uniform smoothing process for only a few pixels near the boundary as in the related art. For example, as described in detail in the examples of FIGS. The smoothing process is performed according to the enlargement ratio.
[0028]
These data are written into the output line FIFO 107 when the data is completely stored in the data storage memory after the enlargement processing.
[0029]
The above is the data output processing after the enlargement processing for the target pixel. Thereafter, this processing is performed while sequentially shifting the target pixel to complete the enlargement processing for one line.
[0030]
As described above, in the image enlargement processing apparatus of the present example, the gradation closer to the original image can be obtained by performing the smoothing process according to the enlargement ratio, instead of performing the uniform smoothing process for several pixels only near the boundary. Enlarge as a held image.
[0031]
The image enlargement processing apparatus of the example shown in FIG. 2 applies linear gradation data as inter-pixel data used in the image smoothing processing in order to efficiently perform the image smoothing processing according to the enlargement ratio. is there.
[0032]
In FIG. 2, reference numeral 201 denotes a line FIFO for storing one line of image data; 202, a post-enlargement storage data storage memory for storing pixels after each enlargement process corresponding to the maximum enlargement ratio; An attention pixel memory for storing and storing data (described as “save pixel of interest” in the figure), 203b is a pixel memory after attention for storing and storing pixel data after the pixel of interest (described as “save pixel of interest” in the figure) , 204 is a pixel counter (described as “pixel number output” in the figure) that counts and outputs the number of pixels for specifying a memory area according to the enlargement ratio, and 205 is an enlargement that holds the enlargement ratio set and instructed by the CPU. A ratio setting register 206 is an inter-pixel interpolation data processing unit that performs an image enlargement process by inter-pixel interpolation according to the present invention.
[0033]
The inter-pixel interpolation data processing unit 206 of this example includes a difference calculator 206a, a magnitude comparator 206b, a divider 206c, a multiplier 206d, and an enlarged interpolation data calculator 206e.
[0034]
The difference calculator 206a calculates a difference between the target pixel data (“A”) stored in the target pixel memory 203a and the pixel data (“B”) after the target pixel stored in the target pixel memory 203b. , The magnitude comparator 206b determines the magnitude of the pixel data of interest (“A”) stored in the pixel memory of interest 203a and the pixel data (“B”) after the pixel of interest stored in the pixel memory of interest 203b ( A> B ?, B> A?).
[0035]
The divider 206c calculates an interpolation count (Δ = δ / n) according to the enlargement ratio using the calculation result (δ) of the difference calculator 206a and the pixel count result (n) of the pixel counter 204. The pixel count result “n” of the pixel counter 204 is used as position information of each piece of interpolation data between the target pixel data and the target pixel data.
[0036]
The multiplier 206d sequentially divides the count number (1 to n) for each pixel count by the pixel counter 204 according to the determination result (A> B ?, B> A?) Of the magnitude comparator 206b. The calculation result (Δ) of 206c is multiplied. For example, if “B> A”, “Δ × 1, Δ × 2,..., Δ × n” is calculated, and if “A> B”, “Δ × n, Δ × n− ,..., Δx2, Δx1 ”.
[0037]
In the post-enlargement interpolation data calculator 206e, based on the determination result (A> B ?, B> A?) In the magnitude comparator 206b and the calculation result in the multiplier 206d, each of the data between the target pixel data and the post-target pixel data is calculated. The values (P1 to Pn) of the pixels (interpolation data) are calculated. For example, if “B> A”, “P1 = A + Δ × 1, P2 = A + Δ × 2,..., Pn = A + Δ × n” is calculated, and if “A> B”, “P1 = B + Δ × n, P2 = B + Δ × n _-1 , ..., Pn _-1 = B + Δ × 2, Pn = B + Δ × 1 ”.
[0038]
As described above, in the multiplier 206c, the Δ (= δ / n) value adjusted to the enlargement ratio (n + 1) is calculated from the difference value (δ) between the pixel data of interest and the pixel data of interest obtained by the difference calculator 206a. .., N in the multiplier 206d by using the Δ value and the positional information n between the pixel data of interest and the subsequent pixel data, and the interpolation data calculator 206e after enlargement. , N) is added to the original image data, and the result is stored at the position (1 to n) of the interpolation data in the post-enlargement storage data storage memory 202.
[0039]
As described above, the inter-pixel interpolation data processing unit 206 in the image enlargement processing apparatus of the example shown in FIG. 2 includes the difference calculator 206b that calculates the difference between the value of the pixel of interest and the value of the pixel after the pixel of interest. A divider 206c that divides the difference value obtained by the operation unit 206b by the enlargement ratio, a multiplier 206d that changes the value of the division result of the divider 206c according to the position of each interpolation pixel, and a multiplier 206d that changes It is composed of an enlarged interpolation data calculator 206e that calculates the value of each interpolation pixel at the position using the calculated value, and a size comparator 206b that compares the size of the pixel data of interest with the pixel data after the pixel of interest. .
[0040]
Hereinafter, an image smoothing process operation according to an enlargement ratio by the image enlargement processing apparatus of the present example including the inter-pixel interpolation data processing unit 206 having such a configuration will be described.
[0041]
The image enlargement processing apparatus of the present embodiment first sets the enlargement ratio in each of the main scanning and sub-scanning directions in each setting register 205. Then, one line of image data read from a scanner or the like is loaded into the line FIFO 201.
[0042]
When reading the image data from the line FIFO 101, a total of two pixels, the enlarged target pixel and the subsequent pixels, are read and stored in the target pixel memory 203a and the target pixel memory 203b, respectively.
[0043]
Based on the pixel data of interest (A) stored in this way and the subsequent pixel data (B), the magnitude comparator 206b compares the magnitudes (A> B ?, B> A?) And obtains the magnitude comparison result. Based on this, the difference calculator 206a calculates the difference (δ) from the larger value to the smaller value (δ = AB, δ = BA).
[0044]
In addition, the number of continuous pixels (n) after the enlargement, that is, the target pixel data stored in the target pixel memory 203a and the post-target pixel memory 203b are determined by the pixel count 204 in accordance with the enlargement ratio set in the setting register 205. The number (n) of pixels between the pixel data after the attention stored in is calculated. In this example, the number of pixels (n) is used as a variable that changes from “1” to “n” as a maximum value, and calculation is performed between the target pixel data and the subsequent pixel data according to the enlargement ratio.
[0045]
That is, the multiplier 206c divides the difference (δ = AB, δ = BA) obtained by the difference calculator 206a by the number of pixels (n) calculated by the pixel count 204, thereby obtaining the pixel data of interest. And a coefficient (Δ = (AB) / n, Δ = (BA) / n) for interpolation calculation with respect to the pixel number (n) between the pixel data and the subsequent pixel data.
[0046]
For example, when the pixel value (B) after the target pixel is larger than the target pixel value (A) (B> A), the multiplier 206d calculates “Δ value × (1, 2,...)” As interpolation data. , N) ”, and the enlarged interpolation data calculator 206e sequentially adds the values obtained by the multiplier 206d using the pixel value of interest (A) as a base value (A + Δ × 1, A + Δ × 2). ,..., A + Δ × n) are sequentially stored in the storage data storage memory 202 after the enlargement processing.
[0047]
Conversely, when the pixel value (B) after the target pixel is smaller than the target pixel value (A) (A> B), the multiplier 206d calculates “Δ value × (n,. 2,1) ”, and the interpolated interpolation data calculator 206e sequentially adds the values obtained by the multiplier 206d using the pixel value (B) after the target pixel as a base value (B + Δ × n,..., B + Δ × 2, B + Δ × 1).
[0048]
As described above, when the pixel value (B) after the target pixel is smaller than the target pixel value (A) (A> B), the pixel value (B) after the target pixel is larger than the target pixel value (A). ) Is larger (B> A), the pixel side after the target pixel is “1” and the target pixel side is “n”.
[0049]
The linear gradation data between the target pixel data and the subsequent pixel data calculated as described above is stored and stored as post-enlargement storage data, which is calculated and determined according to the enlargement ratio set in the setting register 205 by the pixel count 204. The data is stored in the memory area of the memory 102. In this manner, smoothing of the enlarged image data is performed.
[0050]
The image enlargement processing apparatus shown in FIG. 3 applies oversampling filter processing to the image smoothing processing. Hereinafter, the image enlargement processing apparatus shown in FIG. 3 will be described.
[0051]
In FIG. 3, reference numeral 301 denotes a line FIFO for storing one line of image data input from the scanner, and 302 denotes a post-enlargement storage data storage memory for storing pixels after each enlargement process corresponding to the maximum enlargement ratio. , 303a is a pixel memory of interest that stores and stores the pixel data of interest (described as “pixel of interest storage” in the figure), and 303b is a pixel memory of interest that stores and stores pixel data after the pixel of interest (“pixel of pixel after attention” in the figure) ), A pixel counter (described as “pixel number output” in the figure) for counting and outputting the number of pixels for designating a memory area according to the enlargement ratio, and 305, an enlargement designated by the CPU. An enlargement ratio setting register 306 that holds the ratio is an inter-pixel interpolation data processing unit that performs an image enlargement process by inter-pixel interpolation according to the present invention.
[0052]
The inter-pixel interpolation data processing unit 306 of this example has a zero data inserter (described as “0” data embedding unit) 306a, an oversampling filter 306b, and a filter coefficient ROM 306c.
[0053]
The zero data inserter 306a sets “0” between the target pixel (“A”) stored in the target pixel memory 303a and the pixel after the target pixel (“B”) stored in the target pixel memory 303b. Insert At this time, the zero data inserter 306a inserts “0” for the count number (n) counted by the pixel counter 304.
[0054]
An interpolation filter coefficient corresponding to the number of pixels between the target pixel and the subsequent pixels is set in the filter coefficient ROM 306c, and the oversampling filter 306b sets the target pixel (“A”) and the pixel (“A”) after the target pixel. The data sum (A, 0,..., 0, B) composed of “B”) and “0” inserted therebetween is calculated based on the coefficient preset in the filter coefficient ROM 306c. The filter operation is performed by performing the oversampling process by the operation.
[0055]
As described above, the inter-pixel interpolation data processing unit 306 in the image enlargement processing device of the example illustrated in FIG. 3 includes the zero data inserter 306a that inserts zero data into each interpolated pixel between the target pixel and the adjacent pixel, Oversampling filter processing by sum-of-products operation of the value of, the value of the adjacent pixel, each zero data between the pixel of interest and the adjacent pixel, and a coefficient predetermined in the filter coefficient ROM 306c corresponding to the number of interpolation pixels. And an oversampling filter 306c.
[0056]
Hereinafter, an image smoothing process operation according to an enlargement ratio by the image enlargement processing apparatus of the present example including the inter-pixel interpolation data processing unit 306 having such a configuration will be described.
[0057]
First, the image enlargement processing apparatus of this example sets the enlargement ratio in each of the main scanning and sub-scanning directions in each setting register 305. Then, one line of image data read from a scanner or the like is loaded into the line FIFO 301.
[0058]
When image data is read from the line FIFO 301, a total of two pixels, an enlarged target pixel and subsequent pixels, are read and stored in the target pixel memory 303a and the target pixel memory 303b, respectively.
[0059]
In addition, the number of consecutive pixels (n) after the enlargement, that is, the attention pixel data stored in the attention pixel memory 303a and the attention pixel memory 303b, according to the enlargement ratio set in the setting register 305 by the pixel count 304. Calculates the number of pixels (n) between the post-attention pixel data stored in the storage area 308 and a memory area in the post-enlargement storage data storage memory 308 necessary to perform an oversampling filter operation on the number of pixels (n). To secure.
[0060]
Note that the target pixel data stored and stored in the target pixel memory 303a is stored at the head address in the memory area of the post-enlargement storage data storage memory 308, and the last address is stored and stored in the target pixel memory 303b. Saves pixel data after attention.
[0061]
In addition, the zero data inserter 306a stores the value “0” in association with areas other than the first address and the last address in the memory area secured in the post-expansion storage data storage memory 308.
[0062]
Then, in the oversampling filter 306c, a predetermined coefficient is selected from the interpolation filter coefficient ROM 306b corresponding to the number of pixels (n) calculated by the pixel count 204, and the selected coefficient and the post-enlargement storage data storage memory 308 -Sum operation of the pixel data of the first address (A) and the pixel data of the last address (B) in the memory area secured in the above, and the pixel data (0) of the other area embedded by the zero data inserter 306a Is performed, an oversampling operation is performed, and the output result is stored in the storage data storage memory 308 after the enlargement processing. Thereafter, the above processing is repeated to smoothen the enlarged image data.
[0063]
The image enlargement processing device shown in FIG. 4 determines whether or not to perform the smoothing process depending on the type of the image with respect to the image smoothing process, thereby improving the quality of the enlarged image. Hereinafter, the image enlargement processing apparatus shown in FIG. 4 will be described.
[0064]
The image enlargement processing apparatus in FIG. 4 determines the type of image as preprocessing for performing interpolation image data processing for smoothing enlarged image data according to the enlargement ratio described in the examples of FIGS. 1 to 3. Things.
[0065]
The image enlargement processing apparatus of the present example includes n line FIFOs (0) 401a to (n) 401b for storing image data of one line each input from the scanner, and each line corresponding to the maximum enlargement ratio. A post-enlargement storage data storage memory 402 for storing pixels after the enlargement processing, a determination image area storage memory 403 for storing image data to be processed among the image data stored in the n line FIFOs 401a to 401b, FIG. 3 to 406, which perform interpolation image data processing for smoothing the enlarged image data according to the magnification shown in FIG. 3 (described as "enlarged image smoothing processing block" in the figure) 406, It has an output FIFO 407 for outputting the subsequent image data, and has an edge determination unit (described as “edge determination” 412 in the figure) 412 for determining the type of the image. A determination unit (described as “isolated point determination” in the figure) 413, a character image determination unit (described as “character image determination” in the figure) 414, and an image repetition process for simply enlarging image data in accordance with an enlargement ratio (Referred to as “image repetition processing block” in the figure) 411 and selection for selecting one of the processing results of the enlarged image smoothing processing unit 406 and the image repetition processing unit 411 based on the determination result of the character image determination unit 414 415 is provided.
[0066]
As described above, the image enlargement processing device of the example shown in FIG. 4 includes the character image determination unit 414 that determines whether the image data to be enlarged is a character image, and the character image determination unit 414 If not, the selector 415 for selecting and executing the image smoothing process by the enlarged image smoothing processing unit (inter-pixel interpolation data processing unit) 406 is provided. A section 414 performs an edge determination on the image data to be enlarged and an isolated point determination section 413 that performs an isolated point determination on the image data to be enlarged based on the determination results of the isolated image. It is necessary to enhance the edge of a character image by deciding whether or not the data is a character image, by deciding whether or not to perform smooth processing depending on the type of the image. Without performing the smooth processing for the still picture to make it possible to perform enlargement as a more clear image.
[0067]
With such a configuration, the image enlargement processing device of the example shown in FIG. 4 controls the execution of the image smoothing process according to the enlargement ratio as described below.
[0068]
That is, the image enlargement processing apparatus of the present embodiment first writes image data in line FIFOs (0) 401a to (n) 401b for storing image data of a plurality of lines for image determination in an arbitrary image area. . The n line FIFOs (0) 401a to (n) 401b are provided for storing image data for sub-scanning direction processing.
[0069]
Image data for each of the front, rear, upper and lower determination areas for the target pixel is read from the line FIFOs (0) 401a to (n) 401b into the determination image area storage memory 403, and the edge determination unit 412 performs edge determination on the read image data. The isolated point determination unit 413 performs an isolated point determination, and the character image determination unit 414 determines whether the image data is a character based on each determination result.
[0070]
When the result of the determination by the character image determination unit 414 recognizes that the image data is a part of the character, the selector 415 does not perform the smoothing process on the enlarged image data by the enlarged image smoothing process 406, The processing is selected so that enlargement in simple pixel data repetition by the image repetition processing unit 411 is performed.
[0071]
When the character image determination unit 414 determines that the image data is other than a character in the determination result, the selector 415 determines that the image repetition processing unit 411 performs enlarged image smoothing instead of simple pixel data repetition. The processing is selected so as to perform the smoothing processing on the enlarged image data by the conversion processing 406.
[0072]
The enlarged image data by the image repetition processing unit 411 selected by the selector 415 or the enlarged image data by the enlarged image smoothing processing 406 is stored in the post-enlargement processing storage data storage memory 402, and then, via the output FIFO 407. Output.
[0073]
By the process selection according to the image type described above, the character image is output as a clear image without blurring the edge portion.
[0074]
As described above with reference to FIGS. 1 to 4, in the image enlargement processing apparatus of the present example, an image processing apparatus having a scaling function for enlarging image data stores image data for one line. A line FIFO, a storage memory for each post-enlargement pixel corresponding to the maximum enlargement ratio, a memory for holding the preceding and succeeding pixel data in the target pixel data, a pixel counter corresponding to the enlargement ratio, and the number of pixels according to the enlargement ratio And a counter that counts the number of images to perform an image smoothing process in accordance with the enlargement ratio. With this configuration, by performing the image smoothing process in accordance with the enlargement ratio, it is possible to uniformly smooth the image at any enlargement ratio.
[0075]
In particular, as shown in FIG. 2, a memory for holding the subsequent pixel data in the target pixel data, a difference calculator for calculating a difference between the target pixel data and the subsequent pixel data, and a comparison for determining the magnitude of the difference value And a calculator (divider) for calculating a Δ value corresponding to the enlargement ratio from the difference value, and a multiplier and an adder for adding the Δ value to the original image data. The smoothing process for the enlarged image is efficiently performed by applying the gradation data.
[0076]
Further, as shown in FIG. 3, a memory for holding the pixel data after the pixel data of interest and a pixel number calculator for calculating the required number of pixels between the pixel data of interest and the subsequent pixel according to the enlargement ratio A “0” data inserter for embedding “0” between the pixel data of interest and a subsequent pixel, and a storage device storing interpolation filter coefficients corresponding to the number of pixels between the pixel data of interest and the subsequent pixel ( A filter ROM) and a product-sum calculator for performing an oversampling filter operation are provided, and oversampling filter processing is applied as inter-pixel data. As described above, by applying the oversampling filter processing as inter-pixel data, efficient smoothing of an enlarged image can be performed.
[0077]
Also, as shown in FIG. 4, a line FIFO for storing image data for sub-scanning direction processing, a memory for holding pixel data before and after the pixel data of interest, and a decision unit for detecting a character image And a selector that controls to perform image enlargement by simply repeating processing without performing the smoothing processing on the character image, and determines whether to perform the image smoothing processing depending on the image type. Make a decision. As described above, by determining whether or not to perform the image smoothing process depending on the image type, it is possible to retain an edge component particularly for a character image with respect to enlargement and output a clear image.
[0078]
It should be noted that the present invention is not limited to the example described with reference to FIGS. 1 to 4 and can be variously modified without departing from the gist thereof. For example, in the example of FIG. 4, the selector 415 selects one of the processing results of the enlarged image smoothing processing unit 406 and the image repetition processing unit 411 based on the determination result of the character image determination unit 414. However, the selector 415 may be configured to select and execute one of the enlarged image smoothing processing unit 406 and the image repetition processing unit 411. In this case, based on the selection result, there is no need to execute any of the processes of the enlarged image smoothing processing unit 406 and the image repetition processing unit 411, and the efficiency of the processing can be further improved.
[0079]
【The invention's effect】
According to the present invention, instead of performing uniform smoothing processing for only a few pixels near the boundary, an image smoothing process is performed in accordance with an enlargement ratio, and an image is uniformly smoothed at any enlargement ratio. It is possible to enlarge an image having a gradation closer to that of the original image. In particular, by applying linear gradation data as inter-pixel data, or by applying oversampling filter processing as inter-pixel data, it is possible to output a smooth image with respect to enlargement. In addition, by determining whether or not to perform the image smoothing process depending on the type of image, it is possible to hold an edge component particularly for a character image and output a clear enlarged image.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a first configuration example of an image enlargement processing device according to the present invention.
FIG. 2 is a block diagram showing a second configuration example of the image enlargement processing device according to the present invention.
FIG. 3 is a block diagram showing a third configuration example of the image enlargement processing device according to the present invention.
FIG. 4 is a block diagram showing a fourth configuration example of the image enlargement processing device according to the present invention.
[Explanation of symbols]
101: line FIFO, 102: storage data storage memory after enlargement processing, 103: pixel storage memory, 104: pixel counter (“pixel counter for specifying memory area”), 105: enlargement ratio setting register, 106: inter-pixel interpolation Data processing unit (“inter-pixel interpolation data processing block”), 107: output line FIFO, 201: line FIFO, 202: storage data storage memory after enlargement processing, 203a: target pixel memory (“target pixel storage”), 203b: Pixel memory after attention (“Save pixel after attention”), 204: pixel counter (“pixel number output”), 205: enlargement ratio setting register, 206: inter-pixel interpolation data processing unit, 206a: difference calculator, 206b: large and small Comparator, 206c: divider, 206d: multiplier, 206e: interpolation data calculator after enlargement, 301: line F FO, 302: storage data storage memory after enlargement processing, 303a: target pixel memory (“preserve target pixel”), 303b: post-target pixel memory (“preserve target pixel”), 304: pixel counter (“output pixel number”) ), 305: enlargement ratio setting register, 306: inter-pixel interpolation data processing unit, 306a: zero data inserter (“0” data embedding unit), 306b: oversampling filter, 306c: coefficient ROM for filter, 401a to 401b: line FIFO (0) to (n), 402: storage data storage memory after enlargement processing, 403: determination image area storage memory, 406: enlarged image smoothing processing unit (“enlarged image smoothing processing block”), 407 : Output FIFO, 412: edge determination unit (“edge determination”), 413: isolated point determination unit (“isolated point determination”), 41 : Character image determining unit ( "character image determining"), 411: image repetitive processing unit ( "image repeat processing block"), 415: selector.

Claims (5)

An image enlargement processing device for enlarging image data,
Interpolated pixel data processing means for calculating the value of each interpolated pixel to be interpolated between the pixel of interest and adjacent pixels for enlargement in accordance with the respective values and the magnification of the pixel of interest and adjacent pixels,
An image enlargement processing device for performing an image smoothing process according to an enlargement ratio. The image enlargement processing device according to claim 1,
The inter-pixel interpolation data processing means includes:
Difference calculation means for calculating a difference between the value of the pixel of interest and a value of a pixel after the pixel of interest, and division means for dividing the difference value obtained by the difference calculation means by the enlargement ratio;
Calculating means for changing the value of the division result of the dividing means according to the position of each interpolation pixel;
Calculating means for calculating the value of each interpolated pixel at the position using the value changed by the calculating means. The image enlargement processing device according to claim 1,
The inter-pixel interpolation data processing means includes:
Zero data insertion means for inserting zero data into each interpolation pixel between the target pixel and the adjacent pixel,
Oversampling filter processing is performed by a product-sum operation of the value of the pixel of interest, the value of the adjacent pixel, and each zero data between the pixel of interest and the adjacent pixel, and a predetermined coefficient corresponding to the number of the interpolation pixels. And an image enlargement processing device. The image enlargement processing device according to any one of claims 1 to 3, wherein
Determining means for determining whether the image data to be enlarged is a character image,
An image enlargement processing apparatus comprising: a selection unit that selects and executes an image smoothing process by the inter-pixel interpolation data processing unit if the result of the determination is that the image is not a character image. The image enlargement processing device according to claim 4,
The determining means is:
Edge determination means for performing edge determination on image data to be enlarged;
An isolated point determination unit for performing an isolated point determination on the image data to be enlarged; and whether the image data to be enlarged is a character image based on the determination results of the edge determination unit and the isolated point determination unit. An image enlargement processing device for determining whether or not the image is enlarged.

Images