JP2001043363A - System for identifying picture and character and image processor using the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately distinguish a picture and a character mixed in the same bit map image by judging them on the basis of frequency for adjacent pixels to have the same value. SOLUTION: A picture/character identifying part selects an 8×8 pixel block 17 on the upper left side of the bit map image and selects a pixel 1 at the upper left end inside that concerned block. The differences of pixel values between the pixel 1 and a right adjacent pixel 2 or a lower adjacent pixel 9 are respectively calculated and it is judged whether the difference is '0'. The difference of '0' means the values (colors) of adjacent pixels are completely equal. Similar processing is repeated for the other pixels. When the total number of differences of '0' inside the concerned block 17 is greater than a preset threshold value (n), the concerned block 17 is judged as a character and when that total number is smaller than the threshold value (n), the concerned block 17 is judged as a picture. The same processing is repeated for adjacent blocks.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for identifying pictures and characters existing in the same bitmap image, and an image processing for utilizing the identification method, for example, performing compression or printing. Related to the device.

[0002]

2. Description of the Related Art For example, in the case of compressing a bitmap image, a high compression ratio can be obtained by performing irreversible compression, but the image quality deteriorates. By performing lossless compression, the original image quality can be maintained, but the compression ratio is low. In a conventional image compression apparatus, an appropriate compression method is selected according to the use and purpose of use of an image. For example,
Bitmap images distributed on the Internet, etc., have some image quality degradation to reduce the communication load.
It is often compressed by a lossy compression method such as the JPEG baseline.

[0003]

Image elements and characters which are represented by substantially continuous gradation or multicolor, such as a picture (for example, a photograph or a multicolor paint picture), and have sharp contours, such as characters. When a bitmap image that contains image elements with is mixed by irreversible compression, even if the picture has some image quality deterioration, human vision does not respond sensitively to it,
There is a problem that the image quality of characters is deteriorated. Therefore, there is a demand to apply an optimal compression method to each of a picture portion and a character portion of one image separately. Further, there is a case where it is desired to perform another process on the picture and the character in the same image without being limited to the compression. For example, when an image is printed by a digital printer, there are cases where it is desired to perform halftoning processing only on a picture part or to perform edge smoothing processing only on a character part.

Accordingly, an object of the present invention is to accurately distinguish pictures and characters mixed in the same bitmap image.

Another object of the present invention is to compress a bitmap image in which a picture and a character are mixed at a high compression ratio without deteriorating the image quality of the character.

Still another object of the present invention is to distinguish between pictures and characters included in the same bitmap image and selectively perform image processing only on the pictures or characters.

It is still another object of the present invention to print a bitmap image including pictures and characters at high speed and with high image quality.

[0008]

A picture / character identification apparatus according to a first aspect of the present invention divides a bitmap image into a large number of pixel blocks each composed of a plurality of continuous pixels, and divides the bitmap image into each pixel block. Attention means of interest, calculation means for calculating the frequency of adjacent pixels in the pixel block of interest having the same value, and judging whether the pixel block of interest is a picture or a character based on the frequency Judgment means.

According to this device, a block having a high frequency of consecutive pixels having the same value (the same color) is determined to be a character, and a block having a high frequency of adjacent pixels having a different color is determined to be a picture. This makes it possible to satisfactorily distinguish between characters and diagrams drawn continuously in the same color and pictures such as photographs drawn in substantially continuous gradation or multicolor.

An image processing apparatus according to a second aspect of the present invention performs processing by distinguishing between a bitmap image portion determined to be a picture and a bitmap image portion determined to be a character by the picture / character identification device. . That is, the image processing apparatus performs, for example, a specific process only on one of a picture and a character, or performs different processing on a picture and a character.

An image compression apparatus according to a third aspect of the present invention compresses a bitmap image portion determined to be a picture by the picture / character identification apparatus by an irreversible compression method, and generates a bitmap image determined to be a character. The part is compressed by a lossless compression method. In the image compressed by this device, the picture part is lossy-compressed, so the overall compression ratio is high, and the decompressed image does not have any image quality deterioration in the character part, so that the whole image is human. The image is reflected in high image quality.

In a preferred embodiment, the lossy compression method includes JP
EG lossy compression method is used. The JPEG lossy compression method can compress a picture such as a photograph drawn with substantially continuous gradation or multi-color with a high compression ratio and small image quality deterioration. When decompressed, high picture quality can be maintained for both pictures and characters. The JPEG lossy compression method divides a bitmap image into blocks having a size of, for example, 8 × 8 pixels, and performs processing such as discrete cosine transform and nonlinear quantization for each block. -Block division performed for character identification can be shared with block division for compression.

An image decompression device according to a fourth aspect of the present invention receives a compressed image having a picture portion compressed by a lossy compression method and a character portion compressed by a lossless compression method,
Identification means for identifying a picture part and a character part; means for decompressing the picture part of the compressed image into a bitmap image part of the picture by a decompression method corresponding to the lossy compression method; Means for decompressing a character portion in a character into a bitmap image portion of a character by a decompression method corresponding to a lossless compression method, and receiving a bitmap image portion of a picture and a bitmap image portion of a character to complete a bitmap image Means for making the part.

A printing system according to a fifth aspect of the present invention includes a host device that outputs print data, and a printer that receives and prints the print data, and the host device includes the above image compression device and the image compression device. A print data output device for converting a compressed image output from the device into print data and sending the print data to a printer, and the printer is configured based on the image decompression device and the bitmap image output from the image decompression device. A print processing device for performing printing.

According to this printing system, since the compressed image is sent from the host device to the printer, the transmission time is short, and the bitmap image decompressed by the printer is of high image quality without deterioration of character image quality. . Accordingly, high-quality printing can be performed at high speed.

Various processes according to the present invention can be performed by a computer, and a program for that can be installed or loaded into the computer through various media such as a semiconductor memory, a disk-type storage, and a communication line signal. .

[0017]

FIG. 1 shows a configuration of a JPEG compression apparatus according to an embodiment of the present invention.

The JPEG compression apparatus 1 has an input section 3, a picture / character identification section 5, a JPEG lossless compression section 7, a JPEG lossy compression section 9, and an output section 11. The input unit 3 receives the bitmap image data 13 and inputs the bitmap image data 13.
Is a color system (for example, RGB or CMY) image other than the YUV image, color-converts this into a YUV bitmap image, and further converts the YUV bitmap image into a size suitable for JPEG encoding ( (Typically, 8 × 8 pixels).

The picture / character discriminating section 5 sequentially reads the 8 × 8 pixel blocks obtained by dividing the YUV bitmap image, and determines whether each block is a picture or a character.
Here, the “picture” refers to an image element represented by substantially continuous gradation or multicolor, such as a photograph or a multicolor paint picture. On the other hand, "character" refers to an image element having a sharp outline like a character, and also includes a diagram drawn in a single color.

The JPEG lossless compression section 7 receives a block determined to be a character and compresses the block by the JPEG lossless compression method. The JPEG lossy compression unit 9 receives a block determined to be a picture, and compresses the block by a JPEG lossy compression method such as a JPEG baseline.

The output unit 11 converts the character block compressed by the JPEG lossless compression unit 7 and the picture block compressed by the JPEG lossy compression unit 9 into JPEG image data 15 of a predetermined format representing one image. Collectively output. The JPEG image data 15 includes data of a compression result of each block and data necessary for decompressing the compression result of each block (for example, data for identifying whether each block is a picture or a character). And a quantization table used in nonlinear quantization of JPEG lossy compression). The “JPEG image data” referred to in this specification is created by using both the irreversible JPEG compression method and the lossless JPEG compression method as described above. Note that the format is different from the compressed image data (compressed using only one JPEG compression method).

FIG. 2 shows the flow of processing performed by the picture / character identification unit 5.

First, the picture / character identification unit 5 focuses on the upper left 8 × 8 pixel block of the YUV bitmap image (step S1), and focuses on the upper left pixel in the focused block (S2). . FIG. 3 shows one focused 8 × 8 pixel block 17. At first, the first pixel at the upper left corner in the block 17 is focused on.

Next, the process proceeds to step S3, where the difference between the pixel value of the pixel of interest (for example, the first pixel in FIG. 3) and the pixel on the right thereof (for example, the second pixel in FIG. 3) is calculated. Then, it is determined whether or not the difference is 0.
Further, the difference between the pixel value of the focused pixel (for example, the first pixel in FIG. 3) and the pixel below it (for example, the ninth pixel in FIG. 3) is also calculated, and it is determined whether the difference is 0 or not. I do. Here, the “pixel value difference” is YU
In the case of a V image, it means a set of a difference between Y component values, a difference between U component values, and a difference between V component values. Therefore, the fact that the difference is 0 means that the values (colors) of adjacent pixels are exactly the same.

Next, the target pixel is moved to a pixel (for example, the second pixel in FIG. 3) on the right of the current target pixel (for example, the first pixel in FIG. 3) (S5). Step 3 is repeated. Steps S5 and S3 are repeated until the second pixel from the right end of the current row (for example, the 7th pixel in FIG. 3). When the determination processing of S3 is completed for the top row in the block (Yes in S4), next, attention is paid to the leftmost pixel (for example, the ninth pixel in FIG. 3) of the second row from the top next to the bottom row. The pixels are moved, and steps S3 to S5 are repeated for the second row.

Hereinafter, the same processing is repeated up to the seventh row from the top (row starting with the 49th pixel in FIG. 3).
When the determination processing in step S3 is completed for the second pixel from the right end of the seventh row (the fifth pixel in FIG. 3),
The determination as to whether or not adjacent pixels have the same value (same color) is completed for all pixels of the target block (Yes in S8).

Next, proceeding to step S8, it is determined whether or not the total number of differences in the target block that was 0 (ie, the total number of adjacent pixel pairs having the same value) is equal to or greater than a preset threshold n. Judge. As a result, the number of 0 differences becomes the threshold n
If so, it is determined that the block of interest is a "character" (S9). On the other hand, if the number of 0 differences is less than the threshold value n, it is determined that the block of interest is “picture” (S10).

When the first block in the image is determined to be a character or a picture, the block of interest is moved to an adjacent block (S12), and the above processing is repeated to identify a picture or a character. When this identification is completed for all blocks in the image (Yes in S11), the operation of the picture / character identification unit 3 is completed.

According to the above-described identification method, a block having a high frequency in which the values (colors) of adjacent pixels are the same is determined to be a character. Conversely, a block having a high frequency in which the values (colors) of adjacent pixels are different is a picture. Is determined. This identification method is particularly advantageous when JPEG lossy compression is used as the lossy compression method. The reason is that JPEG lossy compression is generally evaluated as not very suitable for images with a small number of colors or images with sharp outlines, because image quality degradation is conspicuous, but according to the above-described identification method, This is because such an image to which JPEG lossy compression is not suitable can be identified as characters. In addition, according to the above-described identification method, even in an image having a large number of colors,
If the same color is continuous, it can be identified as a character, so that a character portion can be accurately picked up as a character from an image in which the character is overlaid on a color photograph, for example.

FIG. 4 shows the data compressed by the JPEG compression apparatus 1 described above.
1 shows a configuration of a JPEG decompression device for decompressing a JPEG image into a bitmap image.

The JPEG decompression device 21 has an input unit 23 and a reversible JP
An EG decompression unit 25, an irreversible JPEG decompression unit 27, and an output unit 29 are provided. The input unit inputs the JPEG image data 31, and based on data for identifying whether each block included in the JPEG image data is a picture or a character, the compression result data of each block included therein is subjected to lossy compression. To determine if the data has been losslessly compressed.

The reversible JPEG decompression unit 25 converts a block of characters
The JPEG lossless compression result data is received from the input unit 23, and is expanded into an 8 × 8 pixel block in the reverse procedure of the JPEG lossless compression. The irreversible JPEG decompression unit 27
JPEG lossy compression result data is received from the input unit 23,
This is decompressed into 8 × 8 pixel blocks by the reverse procedure of JPEG lossy compression.

The output unit 29 connects the decompressed character block and the picture block in the same arrangement as before compression to form one YUV
Create a bitmap image and convert the YUV bitmap image to a bitmap image 3 of an appropriate color system such as RGB.
3 and output.

FIG. 5 shows the configuration of a printing system according to an embodiment of the present invention.

A printer 43 that receives the print data and performs printing is connected to a host device 41 such as a personal computer that generates print data. The host device 41 has the JPEG compression device 1 and the print data output device 45 shown in FIG. 1 in, for example, a printer driver program. When the printer driver of the host device 41 creates print data, the JPEG compression device 1 converts the original, for example, RGB bitmap image into a JPEG compressed image.
As described above, the picture portion included in the bitmap image is compressed by the JPEG lossless compression method, and the character portion is compressed by the JPEG lossless compression method. The print data output device 47 converts the JPEG image data created by the JPEG compression device 1 into print data and transmits the print data to the printer 43.

The printer 43 includes a print data interpretation device 47
And the JPEG decompression device 21 shown in FIG.
9 Print data interpreter 47 and JPEG decompressor 2
1 is provided, for example, in firmware of a microcomputer in a printer. When the print data from the host device 41 is received by the printer 43, the print data interpretation device 47 first interprets the print data, and passes the JPEG image data included in the print data to the JPEG decompression device 21. The JPEG decompression device 421 decompresses the JPEG compressed image data and develops it into, for example, an RGB bitmap image (or may convert directly from YUV to CMYK suitable for printing). The print processing device 49 performs printing based on the expanded bitmap image.

According to this print system, the bitmap image data is compressed into a JPEG image and the host device 41
Is transmitted to the printer 43, the transmission time is short. In the JPEG image, the character portion is losslessly compressed. Therefore, when the JPEG image is bitmap expanded by the printer 43, the image quality of the character portion is not degraded, and a high quality print result is obtained.

Although the embodiments of the present invention have been described above, the present invention can be embodied in various other forms without departing from the gist thereof. For example, after identifying pictures and characters, other image compression methods other than JPEG conversion can be applied to pictures and characters, and other image processing such as edge smoothing can be applied to only characters, for example. . The pixel block to be identified as a picture or a character need not be a square 8 × 8 pixel, but may be, for example, a block of rows, columns, bands, or other shapes, and may be a block of the original bitmap image. T-optimal ones can be selected according to the characteristics and the purpose of image processing.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a JPEG compression device according to an embodiment of the present invention.

FIG. 2 is a flowchart showing processing performed by a picture / character identification unit 5;

FIG. 3 is a diagram showing an 8 × 8 pixel block.

FIG. 4 is a block diagram showing a configuration of a JPEG decompression device.

FIG. 5 is a block diagram showing a configuration of a print system according to an embodiment of the present invention.

[Explanation of symbols]

 1 JPEG compression device 5 Picture / character identification unit 7 JPEG lossless compression unit 9 JPEG lossless compression unit 178 8 × 8 pixel block 21 JPEG decompression device 23 Input unit 25 Reversible JPEG decompression unit 27 Lossy JPEG decompression unit 41 Host device 43 Printer 45 Print data output Device 47 Print data interpretation device 49 Print processing device

Claims (20)

[Claims] 1. A bitmap image is divided into a large number of pixel blocks each composed of a plurality of continuous pixels, and an attention unit that pays attention to each pixel block and an adjacent pixel in the pixel block that is attention have the same value. A picture / character discriminating apparatus comprising: calculating means for calculating a frequency having the following; and judging means for judging whether the pixel block of interest is a picture or a character based on the frequency. 2. The calculating means according to claim 1, wherein all pairs of adjacent pixels in the pixel block of interest are
It is checked whether or not the pixels of each pair have the same value, and the number of pairs having the same value is counted. If the number of counted pairs is equal to or more than a predetermined threshold value, 2. The method according to claim 1, wherein the determined pixel block is determined to be a character, and otherwise determined to be a picture.
Picture / character identification device described. 3. The picture / character identification apparatus according to claim 1, a bitmap image portion occupied by a pixel block determined to be a picture by said picture / character identification apparatus, and a bitmap occupied by a pixel block determined to be a character. Means for performing processing while distinguishing between image parts. 4. The picture / character identification device according to claim 1, further comprising: means for compressing a bitmap image portion occupied by a pixel block determined to be a picture by said picture / character identification device by a lossy compression method; Means for compressing a bitmap image portion occupied by a pixel block determined to be a character by the character identification device by a reversible compression method. 5. The image compression apparatus according to claim 4, wherein said irreversible compression method is a JPEG irreversible compression method. 6. When the irreversible compression method divides the bitmap image into compression pixel blocks of a predetermined size and performs a compression process for each compression pixel block, 5. The image compression device according to claim 4, wherein a focus unit of the character identification device divides the bitmap image into the pixel blocks having the same size as the compression pixel blocks. 7. A picture portion compressed by a lossy compression method,
Receiving a compressed image having a character portion compressed by a reversible compression method, identifying means for identifying the picture portion and the character portion; and identifying the picture portion in the compressed image by the non- Means for expanding a bitmap image portion of a picture by a decompression method corresponding to a lossless compression method, and converting the character portion of the compressed image into a bitmap image portion of a character by a decompression method corresponding to the lossless compression method An image decompression device comprising: means for decompressing; and means for receiving a bitmap image portion of the picture and a bitmap image portion of the character to form a complete bitmap image portion. 8. A printing system comprising: a host device that outputs print data; and a printer that receives and prints the print data, wherein the host device comprises: the image compression device according to claim 4; 8. A print data output device for converting a compressed image output from the device into print data and sending the print data to the printer, wherein the printer is an image decompression device according to claim 7, and a bitmap output from the image decompression device. A print system having a print processing device that performs printing based on an image. 9. A host device used in the print system according to claim 8. 10. A printer used in the print system according to claim 8. 11. A computer-readable recording medium carrying a program for causing a computer to function as the host device according to claim 9. 12. A computer-readable recording medium carrying a program for causing a computer to function as the picture / character identification device according to claim 1. 13. A computer-readable recording medium carrying a program for causing a computer to function as the image processing apparatus according to claim 3. 14. A computer-readable recording medium carrying a program for causing a computer to function as the image compression apparatus according to claim 4. 15. A computer-readable recording medium carrying a program for causing a computer to function as the image decompression device according to claim 7. 16. A step of dividing a bitmap image into a number of pixel blocks each composed of a plurality of continuous pixels, focusing on each pixel block, and determining whether adjacent pixels in the focused pixel block have the same value. A picture / character identification method, comprising: calculating a frequency of the pixel block; and determining whether the pixel block of interest is a picture or a character based on the frequency. 17. A step of dividing a bitmap image into a large number of pixel blocks each composed of a plurality of continuous pixels, focusing on each pixel block, and determining whether adjacent pixels in the focused pixel block have the same value. Calculating the frequency of the pixel block; determining whether the pixel block of interest is a picture or a character based on the frequency; a bitmap image portion occupied by the pixel block determined to be a picture; Distinction processing from a bitmap image portion occupied by a pixel block determined to be a character. 18. A step of dividing a bitmap image into a plurality of pixel blocks each composed of a plurality of continuous pixels, focusing on each pixel block, and determining whether adjacent pixels in the focused pixel block have the same value. Calculating the frequency of the pixel block; determining whether the pixel block of interest is a picture or a character based on the frequency; and irreversibly converting a bitmap image portion occupied by the pixel block determined to be the picture. An image compression method comprising: a step of compressing by a compression method; and a step of compressing, by a lossless compression method, a bitmap image portion occupied by the pixel block determined to be a character. 19. Receiving a compressed image having a picture part compressed by a lossy compression method and a character part compressed by a lossless compression method, distinguishing the picture part from the character part Decompressing the picture portion of the compressed image into a bitmap image portion of the picture by a decompression method corresponding to the lossy compression method; and Decompressing into a bitmap image portion of a character by a decompression method corresponding to the lossless compression method; receiving a bitmap image portion of the picture and a bitmap image portion of the character to form a complete bitmap image portion And an image expansion method. 20. A step of distinguishing a bitmap image into a picture portion and a character portion, and using a lossy compression method for the bitmap image portion determined to be the picture, using the bitmap image determined to be the character. For the image portion, using a lossless compression method, compressing the bitmap image, sending the compressed bitmap image to a printer, and expanding the compressed bitmap image in the printer. , Printing step.