JP5742246B2 - Image processing apparatus and program - Google Patents

Description

  The present invention relates to an image processing apparatus and a program.

  The color image data includes formats such as a dot sequential format and a frame sequential format. Here, the dot sequential format refers to a data format in which color components of color image data are switched for each pixel and expressed as single plane data, and the plane sequential format refers to plane data for each color component of color image data. (See, for example, Patent Documents 1 to 3).

  Another format of color image data is a format called index color. Here, the index color refers to a data format that expresses the color of each pixel of an image by an index (reference number) instead of a value indicating the color itself. As a general-purpose file format corresponding to the image representation by index color, for example, there is TIFF (Tagged Image File Format).

Japanese Patent No. 3381827 JP 2000-305891 A JP 2004-220585 A

  An object of the present invention is to convert three-gradation image data in which one surface of an image is expressed by binary data for two surfaces into an index set with a small number of processing times.

The image processing apparatus according to claim 1 of the present invention, the pixels constituting the surface of the image a 3-level image data indicated by any one of three colors, of one plane, each obtained from the surface Acquisition means for acquiring image data expressed by a set of first data and second data which are binary data, and a conversion table for converting a 2N-digit (N ≧ 2) bit string into an N-digit ternary index a reference means for referring to, among the pixels of the acquired first data and the second data was by the acquisition unit, a combination of pixels between which corresponds to the same position of the surface of the image, is referred to by the reference means that by converting each of the one index using the conversion table, a bit string of 2N digits obtained by extracting and linking by N pixels in succession pixel each other, of N digits the index A converting means for converting, the configuration and output means for outputting an index that has been converted by said converting means.

An image processing apparatus according to a second aspect of the present invention is the image processing apparatus according to the first aspect, wherein the conversion table includes a first table that converts ^22N bit sequences into 3 ^N types of intermediate data, and the intermediate table. And a second table for converting data into the index, wherein the conversion means converts the bit string into the intermediate data and then converts it into the index.

According to a third aspect of the present invention, there is provided a program for each pixel constituting a surface of an image on a computer capable of referring to a conversion table for converting a 2N digit (N ≧ 2) bit string into an N digit ternary index. Image data of three gradations representing any one of three colors, each representing image data expressed by a set of first data and second data, which is binary data for one surface obtained from the surface a first step of, among the pixels of the acquired first data and the second data has been in the first step, a combination of pixels between which corresponds to the same position of the surface of the image, by using the conversion table by converting to a single index, respectively, to convert a bit string of 2N digits obtained by extracting and linking by N pixels in succession pixel each other, the index of N-digit A second step, is intended to execute a third step of outputting the converted index in the second step.

According to the configuration of claims 1 and 3, compared to the case of converting three-gradation image data represented by binary data for two surfaces in a pixel unit (for example, one pixel) smaller than the configuration, It is possible to convert to a set of indexes with a smaller number of conversions.
According to the configuration of the second aspect, it is possible to reduce the total number of N-digit indexes as compared with the case where intermediate data is not used.

Block diagram showing the overall configuration of the image processing system Block diagram showing the hardware configuration of the image input device Diagram illustrating conversion table Diagram for explaining conversion rules The figure for explaining the meaning of the bit string which is the input value of the conversion table Functional block diagram showing the functional configuration of the image processing unit Schematic diagram illustrating data conversion by image processing unit The figure which illustrates the 1st table in case of N = 4 The figure which illustrates the 2nd table in case of N = 4 The figure which illustrates the 1st table in case of N = 8 The figure which illustrates the 2nd table in case of N = 8

[First Embodiment]
FIG. 1 is a block diagram showing the overall configuration of the image processing system according to the first embodiment of the present invention. An image processing system 10 shown in FIG. 1 includes an image input device 100 and an image output device 200.

  The image input apparatus 100 reads an original document that is the source of image data, generates image data, and supplies the image data to the image output apparatus 200. The image input device 100 is, for example, a so-called scanner. The image output device 200 outputs image data supplied from the image input device 100. In the present embodiment, the output by the image output device 200 corresponds to visualizing image data. The image output device 200 is, for example, a display device that displays an image according to image data, or an image forming device that forms such an image on a recording medium such as paper, and more specifically, a personal computer or printer. It corresponds to. Alternatively, the image processing system 10 may be a copier or a facsimile machine that has both a scanner function and a printer function.

  In the present embodiment, it is assumed that the document is expressed by three gradations of red, black, and white. Therefore, the image data of this embodiment (image data supplied by the image input apparatus 100) is expressed in three colors of red, black, and white, and these three colors are never mixed. That is, the image data supplied by the image input apparatus 100 and handled by the image output apparatus 200 is ternary image data.

  FIG. 2 is a block diagram illustrating a hardware configuration of the image input apparatus 100. As shown in FIG. 2, the configuration of the image input apparatus 100 is roughly divided into an image reading unit 110, an image processing unit 120, a storage unit 130, and a communication unit 140. The image processing unit 120 corresponds to an example of the image processing apparatus of the present invention, and has a functional configuration described later.

  The image reading unit 110 is a unit that optically reads an image as a document, generates binary data, and supplies this as image data. The image reading unit 110 expresses three gradations by generating two planes of binary data in the frame sequential format for the same plane. The two binary data generated by the image reading unit 110 are data expressed by binary pixels in which a surface of a certain document is arranged in a plurality of rows and a plurality of columns, one of which is the black position of the document and the other is not. The position is represented by “1 (black)” and “0”, and the other position is represented by “1 (red)” and “0”. In these two binary data, a position that is not black or red, that is, a position corresponding to a pixel indicating “0” in any binary data corresponds to white.

  In the following, for convenience of explanation, binary data describing the black position of the document is referred to as “black and white data”, and binary data describing the red position of the document is referred to as “red and white data”. The image data supplied from the image reading unit 110 to the image processing unit 120 is data configured by combining such black and white data and red and white data. The black and white data and the red and white data correspond to examples of the first data and the second data of the present invention.

  In order to generate such image data, the image reading unit 110 may separately include an image sensor for generating black and white data and an image sensor for generating red and white data, for example, two scans. Thus, both black and white data and red and white data may be generated by a single image sensor. Further, the black and white data and the red and white data may be recognized by the image processing unit 120 as binary data in the frame sequential format, and from the image reading unit 110 in the frame sequential format. There is no need to be supplied to the image processing unit 120. For example, the image reading unit 110 may perform serial communication with the image processing unit 120 and alternately transmit black and white data and red and white data part by part.

  The image processing unit 120 is means for converting the image data supplied from the image reading unit 110 (that is, a set of black and white data and red and white data) into a set of indexes corresponding to the index colors. The image processing unit 120 includes an arithmetic processing device that performs arithmetic processing such as bit arithmetic, and a rewritable image memory for temporarily storing black and white data and red and white data. The image processing unit 120 obtains and outputs one index per pixel by using a conversion table for converting a bit string into an index. In this embodiment, the index corresponding to white is “0”, the index corresponding to red is “1”, and the index corresponding to black is “2”. However, which value of the index is assigned to which color is not limited to this example.

  The storage unit 130 is a unit that stores a conversion table referred to by the image processing unit 120. The conversion table is a data group in which a 2N-digit bit string is stored in association with an N-digit ternary index set. The value of N is an integer equal to or greater than 2, and is desirably a multiple of 4 or 8, but the optimum value depends on the hardware performance of the image input apparatus 100. The value of N corresponds to the number of pixels to be processed at once. Note that if the conversion table outputs a set of predetermined N-digit ternary indexes as an output value when a certain 2N-digit bit string is input as an input value, specific data thereof The structure is not particularly limited. Further, the storage unit 130 is sufficient to be configured by a ROM (Read Only Memory), but may be a rewritable memory.

  The communication unit 140 is means for communicating with the image output device 200. The communication unit 140 transmits the set of indexes converted and output by the image processing unit 120 to the image output apparatus 200. As a result, the image output device 200 recognizes the index list as index color image data.

  FIG. 3 is a diagram illustrating a conversion table according to this embodiment. FIG. 4 is a diagram for explaining the conversion rule in the present embodiment. Here, the value of N described above is “4”. That is, the conversion table shown in FIG. 3 is data for converting an 8-digit bit string into a 4-digit index. Note that the conversion rule is an agreement in advance and does not need to be stored as data in the storage unit 130.

  The conversion rule corresponds to a correspondence between a combination of pixels corresponding to a common position in black-and-white data and red-white data and an index. In this embodiment, the combination of black and white data “0” and red and white data “0” corresponds to an index of “0” (ie, white), black and white data “0” and red and white data “1”. The combination of “1” (that is, red) corresponds to the index “1” of black and white data and the combination of “0” of the red and white data corresponds to the index of “2” (that is, black). The combination of “1” and red / white data “1” corresponds to the index “2” (ie, black).

  Note that since the image data of this embodiment has three gradations, a pixel in which both black-and-white data and red-and-white data are “1”, that is, a pixel that is both black and red cannot exist as a color. However, depending on the mode of reading and binarization in the image reading unit 110, such a pixel that cannot exist may occur. Therefore, any predetermined index is assigned to such a pixel. In this example, the combination of “1” of black and white data and “1” of red and white data is regarded as “2”, that is, black, but other indexes may of course be assigned.

  In FIG. 3, an 8-digit bit string as an input value is obtained by concatenating binary data of four consecutive pixels corresponding to a common position of black and white data and red and white data. In the present embodiment, the upper 4 digits of this bit string are black and white data for 4 pixels, and the lower 4 digits are red and white data for 4 pixels.

  FIG. 5 is a diagram for explaining the meaning of the bit string that is the input value of the conversion table. Here, a case where black and white data indicating a value of “01010101010101101...” And red and white data indicating a value of “00111010000110011. In this case, in the bit string from the first pixel to the fourth pixel from the top, black and white data is “0101” and red and white data is “0011”. When these four-digit bit strings are concatenated, they are “01010011”, which is “83” when expressed in decimal. In addition, when the bit string from the fifth pixel to the eighth pixel is expressed in this way, it is “01010100” in binary number and “84” in decimal number.

  On the other hand, when the image data shown in FIG. 5 is represented by an index according to the conversion rule, the first pixel is “0”, that is, white by a combination of “0” and “0”, and the second pixel is “2”. The third pixel is “1” and the fourth pixel is “2”. Therefore, when the image data from the first pixel to the fourth pixel from the top is expressed by an index, it becomes a 4-digit numerical value of “0212”. Further, when the image data from the fifth pixel to the eighth pixel is expressed by an index in this manner, “0202” is obtained.

  The conversion table shown in FIG. 4 associates the 8-digit bit string thus obtained with the 4-digit index in advance. Therefore, in this conversion table, the output value “0212” is associated with the input value “01010001” (decimal number “83”), and the input value “01010100” (decimal number “84”) is associated with the input value. On the other hand, an output value “0202” is associated.

In the present embodiment, since the binary data has 8 digits, there are 2 ⁸ (2 to the 8th power) ways, that is, 256 ways in total, which can be assumed as input values. On the other hand, there are 3 ⁴ (3 to the 4th power) patterns in total, that is, 81 patterns, since the three types of indexes have 4 digits, as output values. Therefore, this conversion table may give the same output value for different input values. This is because the index when black and white data is “1” and red and white data is “0” and the black and white data and red and white data are both “1” as indicated by the conversion rule described above. This is because both of the indexes are “2”. For example, when the input value is “010100010 (decimal number“ 82 ”)” and “0101010011 (decimal number“ 83 ”)”, the output values of the fourth pixel are both “2”. Therefore, both are “0212”.

  FIG. 6 is a functional block diagram illustrating a functional configuration of the image processing unit 120. The image processing unit 120 includes an acquisition unit 121, a reference unit 122, a conversion unit 123, and an output unit 124. The image processing unit 120 uses the conversion table shown in FIG. 4 as follows, and outputs an index based on image data represented by black and white data and red and white data. The acquisition unit 121, the reference unit 122, the conversion unit 123, and the output unit 124 correspond to examples of the acquisition unit, the reference unit, the conversion unit, and the output unit of the present invention, respectively.

  First, the acquisition unit 121 acquires image data represented by black and white data and red and white data. In the present embodiment, the acquisition unit 121 acquires image data from the image reading unit 110. The reference unit 122 refers to the conversion table stored in the storage unit 130, and obtains an output value (that is, an index) corresponding to a given input value (that is, a bit string).

  The conversion unit 123 extracts a bit string for four consecutive pixels corresponding to a common position from the black-and-white data and the red-white data acquired by the acquisition unit 121 and concatenates them to form an 8-digit bit string. The concatenation of bit strings is realized, for example, by a bit operation in which one is shifted by 4 bits with respect to the other and added. The conversion unit 123 gives the 8-digit bit string to the reference unit 122 as an input value, and obtains an output value corresponding to the input value from the reference unit 122. Thereby, black and white data and red and white data for four pixels are converted into an index for four pixels. The output unit 124 outputs the output value obtained by the conversion unit 123, that is, the index, to the communication unit 140.

  The image processing unit 120 repeats these operations until all of the acquired black and white data and red and white data are processed. At this time, the image processing unit 120 converts the binary data of the four pixel groups from the first pixel to the fourth pixel into the index, and then the binary data of the four pixel groups from the fifth pixel to the eighth pixel. The bit strings corresponding to adjacent (ie, consecutive) pixel groups are sequentially processed from the beginning of the line, such as converting the image into an index, and when the end of the line is reached, these processes are performed from the beginning of the next line. Repeat in order.

  FIG. 7 is a schematic view illustrating data conversion by the image processing unit 120. As shown in FIG. 7, the image processing unit 120 outputs image data represented by a ternary index from black and white data and red and white data that are binary data for two surfaces. The image data output by the image processing unit 120 may be general-purpose 8-bit (that is, 256 colors) index color image data. However, even in this case, only three types of values (that is, three values) are actually used as indexes, and the remaining 253 types of indexes are not used as values indicating pixel gradation.

  Note that the image input apparatus 100 may transmit the image data as TIFF data to the image output apparatus 200, or may transmit it as data of another format. Alternatively, the image output device 200 may have a function of converting image data transmitted from the image input device 100 into TIFF data.

[Second Embodiment]
The second embodiment of the present invention described below is a modification of the configuration of the first embodiment described above, particularly related to the conversion table. Since the other configuration of the present embodiment is similar to that of the first embodiment, the overlapping description is omitted.

  The conversion table of the present embodiment is expressed by two tables (hereinafter, referred to as “first table” and “second table”) using intermediate data. That is, in this embodiment, after the 2N-digit bit string is once converted into intermediate data, the intermediate data is further converted into an N-digit index. The intermediate data is a value given for convenience and may be any value as long as the values are different from each other, but it is desirable that the intermediate data have some regularity with respect to the arrangement of the bit string or the index. Here, for convenience of explanation, the intermediate data is expressed in decimal numbers.

8 and 9 are diagrams illustrating the first table and the second table when N = 4, respectively. In order to easily understand the contents of the conversion table of the present embodiment, the input values illustrated in FIG. 8 are the same as the input values illustrated in FIG. In this example, the first table shown in FIG. 8 is a table for converting the intermediate data of 8-bit sequence of 81 kinds of 256 (2 ⁸ ways) (3 ⁴ kinds). The second table shown in FIG. 9 is a table for converting 81 types of intermediate data into a 4-digit index.

  The second table is different from the conversion table of the first embodiment in that there is no overlap in output values. Therefore, the total number of values stored as output values in the second table is 81, which is smaller than that (256) in the conversion table of the first embodiment. The total number of output values in the second table matches the total number of combinations that can be assumed for the three types of four-digit indexes.

The first table is set so that the output values of the second table do not overlap. That is, the first table are those input values for the second table (i.e., the output value of the first table) sets the intermediate data so that a total of 81 types (3 ^N Street). In this example, the intermediate data is obtained by performing so-called “reverse lookup” from the output value of the second table. That is, the intermediate data is obtained by sequentially assigning values “0” to “80” to these based on 81 combinations of indexes (“0000” to “2222”). Here, a 4-digit index is regarded as a ternary number, and a smaller value of intermediate data is assigned to a smaller number.

  As described above, the second table has a smaller total number of output values than the conversion table of the first embodiment. Therefore, when the storage capacity required to store the second table and the storage capacity required to store the conversion table of the first embodiment are compared, the former storage capacity is smaller (however, the entire conversion table In order to compare the storage capacities, the storage capacities of the first table must also be taken into consideration).

  The difference between the storage capacity required for the second table and the storage capacity required for the conversion table of the first embodiment becomes more significant as the value of N increases. Therefore, in this embodiment, when compared with the first embodiment, it can be said that the larger the value of N, that is, the number of pixels to be collectively converted, is, the smaller the storage capacity of the storage unit 130 is. However, the larger the value of N, the better. The optimal value is determined by the hardware performance such as the overall processing speed of the image input apparatus 100 and the capacity of the image memory.

10 and 11 are diagrams illustrating the first table and the second table when N = 8, respectively. In this example, 16-bit sequence, the upper 8 digits correspond to black and white data, 8-digit lower is equivalent to the red white data is a total of 65536 (2 ¹⁶ combinations). In addition, the index of the 8-digit, a total of 6561 kinds ^{(3 eight).} Therefore, there are 6561 kinds of intermediate data in total.

When the conversion table in the case of N = 8 is realized as in the first embodiment, that is, without being divided into the first table and the second table, the combination of the input value and the output value is (the output value is the same position). (Including those that become) 65536 ways. That is, the combination of the input and output values in this case is approximately 10 times (more precisely than it (6561 kinds) of the present ^embodiment, 2 2N / ^{3 N} times, i.e. ⁴ N / ^{3 N} times )

[Other Embodiments]
In the present invention, binary data in a frame sequential format may be used for three or more planes to represent an image having four or more gradations. That is, the present invention can be applied not only when expressing three colors by an index but also when expressing three or more colors. In view of this point, the image processing apparatus according to the present invention is:
Image data of M gradations indicating each pixel constituting the surface of the image in one of M colors (M ≧ 3), and expressing the surface by a binary data set for (M−1) surfaces Acquisition means for acquiring image data;
A reference means for referring to a conversion table for converting a 2N-digit (N ≧ 2) bit string into an N-digit M-value index;
A bit string of N (M−1) digits obtained by consecutively extracting and connecting pixels corresponding to a common position in the plane of the set of binary data acquired by the acquisition unit N pixels at a time, Conversion means for converting to the N-digit index using a conversion table referred to by the reference means;
And an output unit that outputs the index converted by the conversion unit.

[Modification]
Embodiments of the present invention are not limited to the above-described embodiments, and may be, for example, the following modifications. Moreover, these modifications may be implemented in combination as necessary.

  First, the image processing apparatus according to the present invention may be provided not on the image input apparatus side but on the image output apparatus side. In the present invention, colors used for image representation are not limited to the three colors (black, red, and white) described above. For example, according to the present invention, when reading a document whose background color is not white, or when a drawn image does not include a red color, the index does not include white or red. May be included.

  In the present invention, the process may be executed not for the entire surface of the image but for each part of the surface (for example, for each row or for every plurality of rows). For example, the present invention includes three image processing apparatuses, and the original is divided into three parts in the column direction such as an upper part, a central part, and a lower part, and the upper part is formed in the first image processing apparatus. May be executed in parallel by a plurality of devices, such as causing the second image processing device to process the lower part and the third image processing device to process the lower part.

  In the present invention, the functional configuration corresponding to the image processing unit 120 described above can also be realized by cooperation of hardware resources such as a CPU (Central Processing Unit) and a memory and software. That is, the present invention can also be implemented in the form of a program for realizing the image processing apparatus according to the present invention using hardware resources, or a recording medium on which the program is recorded.

DESCRIPTION OF SYMBOLS 10 ... Image processing system 100 ... Image input device 110 ... Image reading part 120 ... Image processing part 121 ... Acquisition part 122 ... Reference part 123 ... Conversion part 124 ... Output part 130 ... Storage part 140 ... Communication unit, 200 ... Image output device

Claims (3)

The first and second data are three-level image data representing each pixel constituting one surface of the image in one of three colors, each of which is binary data for one surface obtained from the surface. Acquisition means for acquiring image data expressed by the set;
Reference means for referring to a conversion table for converting a 2N-digit (N ≧ 2) bit string into an N-digit ternary index;
Among the pixels of the acquired first data and the second data was by the acquisition unit, a combination of pixels between which corresponds to the same position of the surface of the image, respectively, using the conversion table is referred to by the reference means by conversion to one index, and converting means for converting the bit string of 2N digits obtained by extracting and linking by N pixels in succession pixel each other, the index of N digits,
An image processing apparatus comprising: output means for outputting the index converted by the conversion means. The conversion table is
A first table for converting 2 ^2N bit strings into 3 ^N kinds of intermediate data;
A second table for converting the intermediate data into the index;
The image processing apparatus according to claim 1, wherein the conversion unit converts the bit string into the intermediate data and then converts the bit string into the index. To a computer that can refer to a conversion table for converting a 2N-digit (N ≧ 2) bit string into an N-digit ternary index,
The first and second data are three-level image data representing each pixel constituting one surface of the image in one of three colors, each of which is binary data for one surface obtained from the surface. A first step of acquiring image data represented by a set;
Conversion of the pixels of the first data and the second data obtained in the first step, a combination of pixels between which corresponds to the same position of the surface of said image, each one index using said conversion table A second step of converting a 2N-digit bit string obtained by extracting and concatenating the pixels consecutively by N pixels into the N- digit index,
A program for executing the third step of outputting the index converted in the second step.

Images