JP3122481B2 - Image encoding apparatus and image encoding method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image encoding apparatus and an image encoding method for encoding an input image, for example, an image obtained by combining a character image and a line image with a natural image.

[0002]

2. Description of the Related Art Conventionally, as a method of encoding a natural image having continuous tone such as a photograph, for example, an image is divided into small blocks of 8 × 8 pixels, and discrete cosine transform (hereinafter, DC) is performed.
After performing an orthogonal transformation such as T), the coefficient is encoded.

[0003]

However, in the conventional example described above, when encoding an image in which a pattern such as a character or a line drawing (graphics) artificially generated by a computer or the like is combined with a natural image, the background is A sharp edge is generated at the boundary between the natural image and the overwritten character / line image, and the quality of the character / line image is significantly reduced. When quantizing the coefficients after orthogonal transformation,
Since the quantization step width of a higher-order coefficient is increased in accordance with the visual characteristics, distortion of a high-frequency component increases, deterioration (mosquito noise) such as ringing occurs at an edge portion, and coding efficiency decreases. There was also a disadvantage.

[0004] The present invention has been made in view of the above drawbacks, letters and line drawings, etc., I do not want to degrade the quality pixel portion
As for, while maintaining good quality,
The main purpose is to perform efficient encoding . Special
Pixels that do not want to degrade the quality
Generated at least locally with a single pixel value
Is different from the pixel value distribution of the background part when
And efficiently extract the above pixel portion and separately encode it.
For a further purpose. In addition, the pixel value of the above pixel portion is
With the additional purpose of being decrypted without
I do.

[0005]

To achieve the above object, according to the Invention The image encoding apparatus of the present invention, the blanking of the input image data
The frequency of occurrence of pixel values included in the block is detected for each lock.
Out on the occurrence frequency distribution, and
Determining means for determining a pixel corresponding to a pixel value having a frequency as an extracted pixel; and for each pixel of the input image data,
A determination result indicating whether or not the pixel is the extracted pixel is generated.
First encoding means for encoding a result and generating first encoded data;
Replacement means for replacing the corresponding pixel value with another predetermined value;
Encode the input image after replacement and generate second encoded data
Second encoding means for performing the first and second encoded data,
And output means for outputting data indicating the pixel value of the extracted pixel . In addition,
In order to achieve this, the image coding method of the present invention uses the input image data
For each block of data, the generation of pixel values contained in that block
Detect the raw frequency, isolate on the occurrence frequency distribution, and
Pixels corresponding to pixel values having the above occurrence frequencies are extracted
A determining step of determining that the pixels are the same and each pixel of the input image data
In response, a determination result indicating whether the pixel is the extracted pixel is
Generate and encode the result to generate first encoded data.
A first encoding step, and
Replacement that replaces the pixel value corresponding to the extracted pixel with another predetermined value
Encoding the input image after the replacement and the second encoded data.
A second encoding step for generating data, and the first and second codes
And the data indicating the pixel value of the extracted pixel.
And a power output step.

[0006]

[0007]

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A preferred embodiment according to the present invention will be described below in detail with reference to the drawings. <First Embodiment> FIG. 1 is a schematic block diagram showing a configuration of an image coding apparatus according to a first embodiment. In the figure, 1
Is an input terminal for inputting image data from a computer or the like (not shown). A character / line drawing extraction circuit 2 extracts a character / line drawing from image data and outputs gradation data and a determination result, which will be described in detail later. Reference numeral 3 denotes an encoding circuit which encodes the extracted characters and line drawings. Reference numeral 4 denotes an extracted pixel replacement circuit, which replaces a pixel corresponding to the extracted character / line image with a predetermined value. An encoding circuit 5 encodes the replaced image data. 6 is a decoding circuit,
The data encoded by the encoding circuit 3 is decoded. A decoding circuit 7 decodes the data encoded by the encoding circuit 5. A synthesizing circuit 8 synthesizes the data from the decoding circuits 6 and 7 with the original image data.

In the above arrangement, a pixel corresponding to a character / line image synthesized by a computer or the like is extracted from the image data input from the input terminal 1 by a character / line image extraction circuit 2, and gradation data of the extracted pixel is obtained. (Including color data of a character / line drawing) and a determination result indicating whether the pixel is an extracted pixel (a pixel extracted as a character / line drawing) (1 bit for each pixel, "1" for an extracted pixel in the embodiment, and "1" for other pixels) ("0" is output) is input to the encoding circuit 3. The input judgment result is arithmetic coding, Huffman coding, MH, MR, MM
The data is encoded by an encoding circuit 3 having a known entropy encoding means suitable for encoding binary data such as R. Further, the gradation data of the extracted pixel is multiplexed with the determination result code coded by the above-described entropy coding and output to the transmission path.

Next, the pixel data input to the extracted pixel replacement circuit 4 has a value (for example,
Average value, median value, mode value,
Alternatively, the data is replaced with data interpolated from the peripheral pixels of the extracted pixel) and input to the encoding circuit 5. The pixel data after the replacement is obtained by using the color still image coding method AD proposed by JPEG (Joint Photographic Expert Group).
The image data is encoded by an encoding circuit 5 having an encoding means suitable for encoding a multi-valued image such as an adaptive discrete cosine transform (CT), DCT, or DST (discrete sine transform) and output to a transmission path.

After that, the judgment result of the extracted pixel and the gradation data of the extracted pixel transmitted from the encoding circuit 3 are decoded by the decoding circuit 6 and output to the synthesizing circuit 8. The image data transmitted from the encoding circuit 5 is
The data is decoded by the decoding circuit 7 and output to the synthesizing circuit 8. Then, when the result of the determination of the decoded extracted pixel is “1” (that is, the extracted pixel), the gradation data of the extracted pixel is selected and the determination result is “0” ( That is, in the case of a pixel other than the extracted pixel), the image data is selected and restored to the original input image signal.

Next, the detailed configuration of the character / line drawing extraction circuit 2 will be described below with reference to the block diagram shown in FIG.

First, the input image data is cut out by the blocking circuit 10 into small blocks of, for example, 8 × 8 pixels. The histogram generation circuit 11 generates a histogram for each block based on the extracted pixel data, and outputs the result to the extracted pixel determination circuit 12.
On the other hand, the image data input to the delay circuit 13 is output to the extracted pixel determination circuit 12 and the extracted pixel replacement circuit 4 after being delayed by the time required for generating the above-described histogram. Next, the extracted gray level is detected from the result of the histogram input by the extracted pixel determination circuit 12, and the image data input from the delay circuit 13 is compared with the above extracted gray level. As a result, when they match, it is determined that the pixel is an extracted pixel, and “1” is output from the signal line 101 as a determination result. When they do not match, “0” is output.

Next, a method of detecting an extracted gradation in the embodiment will be described below with reference to FIG.

The character / line drawing data added by a computer or the like has no correlation with the background, and therefore, the gradation is discontinuous. Also, the gradation of the added character / line drawing is usually
The same data is obtained in a specific area. On the other hand, natural image data such as a photograph has strong correlation between pixels, and noise is superimposed at the time of input, or the distribution is widened due to shading or the like. Therefore, the added character / line drawing data can be determined by detecting a gradation in which the generated histogram has no distribution and the frequency is equal to or greater than the predetermined threshold Th. For example, when a histogram as shown in FIG. 3 is generated, the gray scale C is isolated with a nearby frequency of “0” and the frequency is equal to or higher than the threshold Th, so that the gray scale C is the extracted gray scale. Becomes Further, although the gray scales A and D are also isolated, they are not determined as the extracted gray scales because the frequency is low and the possibility of the isolated point due to noise or the like is high.

Here, a detailed configuration of the encoding circuit 3 for inputting the above-described determination result and the extracted gradation will be described with reference to a block diagram shown in FIG. As described above, the signal line 101
The input determination result is converted to raster scanning by a rasterization circuit 14 in order to improve encoding efficiency.
It is input to the entropy encoding circuit 15. Then, the data encoded by the entropy encoding circuit 15 is input to the multiplexing circuit 19. Note that the entropy coding circuit 15 is a known coding circuit suitable for coding binary data such as arithmetic coding and Huffman coding.

On the other hand, the extracted gradation data input from the signal line 102 is input to the change point detection circuit 16 and the gradation table 17, respectively. The change point detection circuit 16 controls the gate circuit 18 to open the gate of the gate circuit 18 when the input extracted gradation data is different from the data of the previous block, and to close the gate when they are the same. However,
If there is no extracted pixel, it is regarded as the same. When the inputted gradation data is new, the gradation table 17 registers the gradation data on the table and outputs the gradation data and the corresponding index to the gate circuit 18. Also,
If the input data has already been registered, only the index is output to the gate circuit 18 without registration.

Therefore, only when the extracted gradation data changes in the gate circuit 18, the index (or gradation data) corresponding to the gradation data is input to the multiplexing circuit 19, and the multiplexing circuit 19 Multiplexed with the encoded determination result and transmitted. As a multiplexing method, for example, it is conceivable to alternately transmit the determination result and the gradation data coded every page or every n block lines. In addition, the data may be multiplexed at the time of transmission, or may be stored in separate memories at the time of storage. Normally, the number of gradations of characters, line drawings, and the like used in the same page is very small. Therefore, this configuration can greatly reduce transmission data. Especially for full color data, each of RGB is 8
Since the gradation data is a total of 24 bits, the effect is enormous.

Next, the extracted pixel replacement circuit 4 in the embodiment.
Will be described with reference to the block diagram shown in FIG.

First, the extraction determination result input from the signal line 101 is input to the replacement value generation circuit 20 and the delay circuit 22, and the input image data input from the signal line 103 is input to the replacement value generation circuit 20. Each is input to the delay circuit 21. Here, the replacement value generation circuit 20 is a circuit that generates data that replaces the extracted pixel data. When the replacement image data is coded by the coding circuit 5, the replacement value generation circuit 20 improves the coding efficiency. Occurs. In particular,
A block average value or the like of the pixel data excluding the extracted pixels is generated. The delay circuits 21 and 22 are circuits for delaying the input image data and the extraction determination result by a time required for generating a replacement value.
Is input to The selector 23 determines, according to the extraction determination result,
If the input image data is an extracted pixel, the replacement value generation circuit 24
Otherwise, the image data of the delay circuit 21 is selected, and the data of the extracted pixel portion is replaced.

FIG. 6 is a block diagram showing the structure of the synthesizing circuit 8. The signal line 111 shows the extraction judgment result decoded by the decoding circuit 6, the signal line 112 shows the extracted gradation data, and It is assumed that image data decoded by the decoding circuit 7 is input to the signal line 113.

In the figure, extracted gradation data and image data are converted into raster scans by rasterization circuits 24 and 25, respectively, and at the same time, a pixel at the same position as a pixel of the extraction judgment result input from a signal line 111 is displayed. And the signals are input to the selector 26 together. Selector 26
Then, when the extraction determination result indicates the extracted pixel, the extracted pixel data is synthesized by selecting the extracted gradation data, and otherwise, the image data is selected, and the input image data before replacement is restored.

As described above, according to the first embodiment, it is possible to efficiently compress image data while maintaining the quality of characters and line drawings. <Second Embodiment> Next, a second embodiment according to the present invention will be described in detail with reference to the drawings.

FIG. 7 is a schematic block diagram showing the configuration of an image coding apparatus according to the second embodiment. In the figure, components having the same functions as those in FIG. 1 are denoted by the same reference numerals, and only different points from the first embodiment will be described.

As shown in FIG. 7, this embodiment is a case where the image coding apparatus is applied to a page printer, and the transmission lines 121 and 122 shown in FIG.
7, 28. In this configuration, the data encoded by the encoding circuits 3 and 4 are sequentially stored in the respective memories 27 and 28, and when one page of data is stored, a printer engine (not shown) is activated. Is done. Here, a synchronization signal from the printer engine is input from the input terminal 31, and the decoding circuits 29 and 30 perform decoding in synchronization with the above-described synchronization signal. As a result, the output terminal 9
Outputs an image signal synchronized with the printer engine, and executes printing.

According to the second embodiment, characters are data-compressed by entropy coding, so that fonts are not distorted and character quality is preserved. Also, since data such as characters and line drawings that are not good at orthogonal transformation are removed, the memory 28
Can be greatly reduced.

The present invention may be applied to a system including a plurality of devices such as a scanner, an interface, and a printer, or may be applied to an apparatus including a single device such as a copying machine. Further, it is needless to say that the present invention can be applied to a case where the above is achieved by supplying a program stored in a floppy disk or an ID card to a system or an apparatus.

[0028]

As described above, according to the present invention, a pixel portion, such as a character or a line drawing, whose quality is not desired to be degraded is determined.
While maintaining good quality, it is effective for other pixels.
Encoding can be performed efficiently. In particular, the above quality deteriorates
Pixels that you do not want to be removed are at least
When a part having a single pixel value is used in some places,
Considering that it is different from the pixel value distribution,
Frequency of occurrence of pixel value included in each block for each block
Is detected on the occurrence frequency distribution, and
Pixels corresponding to pixel values having an occurrence frequency are determined as extracted pixels.
Both the judgment result and the data indicating the pixel value of the extracted pixel.
, So that the above pixel part can be extracted efficiently.
And can be separately encoded. In addition, the
Is the pixel of the extracted pixel in each block.
Since it is data indicating a value, each extracted pixel is
Reversible playback .

[Brief description of the drawings]

FIG. 1 is a schematic block diagram illustrating a configuration of an image encoding device according to a first embodiment.

FIG. 2 is a block diagram illustrating a configuration of a character / line drawing extraction circuit.

FIG. 3 is a diagram showing a generated histogram.

FIG. 4 is a block diagram illustrating a configuration of an encoding circuit.

FIG. 5 is a block diagram illustrating a configuration of an extracted pixel replacement circuit.

FIG. 6 is a block diagram illustrating a configuration of a synthesis circuit.

FIG. 7 is a schematic block diagram illustrating a configuration of an image encoding device according to a second embodiment.

[Explanation of symbols]

 2 Character / line drawing extraction circuit 3 Encoding circuit 4 Extracted pixel replacement circuit 5 Encoding circuit 6 Decoding circuit 7 Decoding circuit 8 Synthesis circuit

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-4-63064 (JP, A) Television Technical Report, 14 [29] (1990) Ando, Higashino, Yanaka, p. 25-30 (58) Field surveyed (Int. Cl. ⁷ , DB name) H04N 1/41-1/419 G06T 9/00

Claims (6)

(57) [Claims] 1. A block for each block of input image data.
The frequency of occurrence of pixel values contained in the
Pixels that are isolated on the distribution and have a frequency of occurrence equal to or higher than a predetermined value
Determining means for determining a pixel corresponding to a value as an extracted pixel; and for each pixel of the input image data,
Generates a judgment result indicating whether or not there is , and encodes the result
A first encoding unit for generating first encoded data; and an image corresponding to the extracted pixel in the input image data.
Replacing means for replacing the prime value with another predetermined value; encoding the input image after the substitution to generate second encoded data;
A second encoding unit that generates the first and second encoded data, and a pixel of the extracted pixel.
Output means for outputting data indicating a value . 2. The method according to claim 1, further comprising: extracting the first encoded data with the extracted data.
The image encoding apparatus according to claim 1, further comprising a multiplexing unit that multiplexes data indicating a pixel value of the pixel . 3. A detecting means for detecting a change in a pixel value corresponding to the extracted pixel , wherein the data indicating the pixel value of the extracted pixel is obtained only when a change in the pixel value is detected by the detecting means. The image encoding device according to claim 2 , wherein the image encoding device performs multiplexing. 4. The detecting means determines whether a pixel newly determined as an extracted pixel by the determining means appears.
The image coding apparatus according to claim 3, wherein data indicating a pixel value of a pixel determined last is held. 5. The method according to claim 5, further comprising: indicating a pixel value at which the change is detected.
If the data indicating the pixel value at which the change is detected is not registered in the table, the data is registered in the table, and the data is multiplexed by the multiplexing unit. change
4. The data according to claim 3, wherein, if data indicating the pixel value at which is detected is already registered in the table, an index corresponding to the registered data is multiplexed by the multiplexing means. Image coding device. 6. A block for each block of input image data.
The frequency of occurrence of pixel values contained in the
Pixels that are isolated on the distribution and have a frequency of occurrence equal to or higher than a predetermined value
A determining step of determining a pixel corresponding to a value as an extracted pixel; and for each pixel of the input image data,
Generates a judgment result indicating whether or not there is, and encodes the result
A first encoding step of generating first encoded data; and an image corresponding to the extracted pixel in the input image data.
A replacement step of replacing the prime value with another predetermined value; encoding the input image after the replacement to generate second encoded data;
Generating second encoding step, the first and second encoded data, and a pixel of the extracted pixel
Output step of outputting data indicating a value.
Image encoding method to be used .