JPH05284371A - Image data encoding method - Google Patents

Abstract

PURPOSE:To efficiently encode an achromatic color image by encoding only a component decided to be significant with regard to plural components of a picture element. CONSTITUTION:In an image data encoding part 11, Y, Cr and Cb signals of an image component are subjected to ADCT conversion and encoded, stored in a buffer 13, and also, sent to an encoding/calculating part 14, and in the case of being an achromatic color image, the code quantity is calculated as a prescribed value. A code output part 15 decides whether the code quantity in the code quantity calculating part 14 coincides with the prescribed value or not, regards it as a significant component in the case they do not coincide with each other, and outputs the code held in the code buffer 13. Accordingly, a luminance Y signal is outputted since the code quantity is larger than the prescribed value, but a color signal is decided to be equal to the prescribed value, and not outputted. In such a way, code generation of a color component which does not contribute to a restored image in a monochromatic image is suppressed, the code quantity is reduced without being accompanied with deterioration of the picture quality and the efficiency can be enhanced.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a device for compressing image data, and more specifically, it divides each multi-valued image of an image composed of a plurality of components into blocks composed of a plurality of pixels and The present invention relates to an image data encoding method for a multivalued image to be encoded.

[0002]

2. Description of the Related Art In order to store image data whose amount of information is orders of magnitude greater than numerical data, especially halftone image and color image data, or to transmit at high speed and high quality,
It is necessary to efficiently code the gradation value for each pixel. Adaptive Discrete Cosine Transform (A) is a highly efficient compression method for image data.
DCT). ADCT divides an image into blocks of 8 × 8 pixels, converts the image signal of each block into a coefficient of spatial frequency distribution by a two-dimensional Discrete Cosine Transform, and quantizes it with a threshold value suitable for vision. The quantized coefficient obtained is encoded by a Huffman table obtained statistically.

FIG. 10 is a conventional block diagram of a still image coding apparatus. Prior to encoding, a RGB signal of a color image to be input is a luminance signal Y which is a component suitable for encoding.
And the color conversion unit 1 converts into two types of color components Cr and Cb.
The signals of the three types of components converted into the Y, Cr, and Cb signals by the color conversion unit 1 are stored in the image buffer 2. When three types of signals for one screen are stored in the image buffer 2, the encoding control unit 3 operates the image data encoding unit 4. still,
By the control described below, signals for instructing the output of each component are sequentially added to the image buffer 2.

FIG. 11 is an operation flowchart of the image data encoding unit 4. The image data encoding unit 4 separately encodes the Y, Cr, and Cb signals that are the components described above. First, when the processing is started, I = 1 in step ST1.
To set. This I indicates each component, and I
Is 1, for example, a Y signal, 2 is a Cr signal, and 3 is a Cb signal. After step ST1, the encoding of the I-th component is executed in step ST2. At this time, I is 1
Therefore, the Y signal is encoded. Then step ST3
The control for outputting the code of the first component obtained in step S1 is performed, and the image data encoding unit 4 outputs the code obtained by this control. Subsequently, it is determined in step ST4 whether or not the output component is completed, and when it is not completed (NO), I = I + 1 is set. This causes I to change from 1 to 2. Subsequently, steps ST2 and ST3 described above are executed, and by repeating this, the respective components of the Cr signal and the Cb signal are also encoded. Then, by the determination in step ST4, it is determined that all the components are finished (YES), and the process is finished. The data is encoded by the processing of the image data encoding unit as described above.

The image data encoding unit 4 is composed of, for example, a DCT conversion unit, a quantization unit, a variable length encoding unit, etc., and encodes each signal under the control of the encoding control unit 3 in step ST2. For example, the DCT transform unit transforms the Y signal of the first component output from the image buffer 2 into the coefficient of the spatial frequency distribution by the two-dimensional discrete cosine transform. Then, the quantizer quantizes the component with a threshold adapted to vision,
The variable-length coding unit performs variable-length coding on the last-obtained quantized coefficient using a Huffman table that is statistically obtained. The color image is encoded by executing the Y signal, the Cr signal, and the Cb signal by the above operation.

[0006]

In the above-mentioned conventional encoding apparatus, since priority is given to a color image,
For example, even when encoding an achromatic image, all components are encoded, code data is accumulated, and further transmitted. However, even a color image signal has an achromatic image, and even in such a case, a code for a certain amount of color components is required. For example, a medical image is an achromatic image, and when such an image is read and encoded, a slight color unevenness in the background portion that occurs at the time of reading is also encoded, so that the restored image looks dirty. Had a problem. Also, there is a problem that the data compression rate is poor even though the image is an achromatic image.

It is an object of the present invention to provide a decoding method for efficiently coding an achromatic image without the intervention of an operator while being a color image coding apparatus.

[0008]

According to the present invention, for each block obtained by dividing an original image composed of pixels of a plurality of components into a plurality of blocks, the plurality of gradation values in the block are calculated. This is in the encoding method.

[First Invention] The significance of each of the plurality of components of the pixel is determined in the first step, and only the component determined to be significant in the first step is encoded and output. Since the encoded data is output only for the components determined to be significant, efficient encoding is possible.

[Second Invention] All the plurality of components input in the first step are encoded, and the unnecessary component code is removed in the second step based on the encoded code amount.
The coding is performed for all components, but since the codes of unnecessary components are removed based on the code amount, efficient coding is possible.

(Third invention) A distribution range of each component of the plurality of components to be encoded is selected in the first step, and each component is selected from the selection result obtained in the first step. Is determined in the second step, and only the components determined to be significant in the second step are encoded in the third step.

An original image composed of pixels of a plurality of components to be encoded is a signal composed of RGB, for example, and the step of detecting the distribution range of each component in the first step uses RGB signals of Y, Cr,
It is performed when converting to a Cb signal. Then, the judgment as to whether it is significant or not is judged to be significant when the Cr and Cb signals are outside the predetermined range. In particular, when the above-mentioned original image is achromatic, the Cr and Cb signals are not output with only the Y signal, so that only significant components can be encoded.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the drawings.
FIG. 1 is a processing flowchart of the present invention. The first aspect of the present invention is to determine whether or not a component of an input image is significant, and to perform processing according to the result. In step ST11, component numbers are sequentially designated, in step ST12, it is determined whether or not the component is significant, and when the component is not significant, I = I + 1 is performed in step ST13 to designate the next component. Then, the process is repeated from step ST12 again.

When the I-th component is significant (YES)
In step ST14, the code of the I-th component is output following step ST2 (ST14). And
In step ST15, it is determined whether all the components are finished,
If not completed, the process is repeated from step ST13. When the processing for all the components is completed, all the processing is completed (ST15). The above-mentioned components are image signals such as Y, Cr and Cb signals. If they are black and white signals, the Cr and Cb signals have intermediate values and are constant. From this constant value, when the input signal is a monochrome image, it is determined that only the Y signal, which is a luminance signal, is valid, and the others are invalid, and only valid is output.

In the case of a completely monochrome image such as an image read by a CT scanner, all pixel levels of color components in the conventional encoding technique have a constant value (in the case of 256 gradations, 128 gradations). However, the present invention determines that the two types of color components are not significant and does not output the code. Therefore, the same restored image can be restored as compared with the conventional one. Nevertheless, it can be encoded with a smaller code amount. Especially, it works efficiently on a completely monochrome image such as a CT image.

FIG. 2 is a block diagram of the first embodiment of the present invention. The image data encoding unit 11 is, for example, the ADCT shown in FIG.
The conversion is performed and the Y, Cr and Cb signals are encoded. FIG. 3 is a detailed configuration diagram of the image data encoding unit 11. The Y, Cr, and Cb signals that are input data are sequentially added in screen units. That is, each component unit is added. Then, this component unit is finally ADCT-converted in units of 8 × 8 pixels.

First, the DCT converter 111
The three-dimensional discrete cosine transform (DCT) is performed, and the DCT-transformed data is quantized by the quantizer 112. A quantization threshold value 113 is added to the quantization unit 112, and the quantization unit 112 quantizes the DCT-transformed information of 8 × 8 pixels using this quantization threshold value 113. This quantized information is added to the variable length coding unit 114. A Huffman code table 115 is connected to the variable length coding unit 114, and the information quantized by the quantizing unit 112 is variable length coded using the Huffman code table 115.

In order to operate the image data coding unit 11 as described above, the coding control unit 12 controls the image data coding unit 11 so as to code one component for each component forming a color image. Then, the image data encoding unit 11 stores in the code buffer 13 the code data obtained by performing the ADCT conversion for each component as described above.

The code data output from the image data encoding unit 11 is also output to the code amount calculation unit 14 in parallel with the storage in the code buffer 13. Here, in the case of a completely achromatic image, all of the two types of pixel levels of the luminance Y signal, the color Cr signal, and the Cb signal, which are the color components Cr and Cb signals, are halftone. For example, if there are 256 gradations, the pixel level will be 128.

Therefore, all encoded blocks are 0
From the DC difference value and the code indicating the end of the block (EO
B). Since the code for each event is assigned in advance, for example, the DC difference value 0 is 2 bits,
If the EOB is 3 bits, the code amount of each color component is a constant value of (5 × the number of blocks) bits.

The code output unit 15 determines whether or not the code amount in the code amount calculation unit 14 matches a predetermined value, and if the code amount and the predetermined value match, the code output unit 15 outputs the code to the coding control unit 16. Notify the completion of.

When the code amount and the predetermined value do not match, it is regarded as a significant component and the code output unit 15 is instructed to output the code held in the code buffer. As a result, the code output unit 15 outputs the code.

FIG. 4 is an operation flowchart of the above-described first embodiment of the present invention. The code output control unit 16 performs the operation control of FIG. 4 to control the entire operation. First, the code output control unit 16 sets I = 1 in step ST31. Then, the coding of the I-th component is instructed in step ST32. Further, in step ST33, it is determined whether the code amount of the I-th component is larger than a predetermined value. If it is determined in step ST33 that the number is larger than the predetermined value (YES), the sign of the first component is significant, and therefore step ST
At 4, the code of the first component is output.

Subsequently, it is determined in step ST35 whether or not all the components are finished, and when all the components are not finished or when it is not determined in step ST33 that the code amount of the I-th component is larger than a predetermined value. Step NO3 for (NO)
At 6, I = I + 1. The step ST36 indicates the next component.

Subsequently, the set I component is encoded in step ST32, and it is again determined in step ST33 whether the code amount of the I component is larger than a predetermined amount. By repeating the above-described operation in sequence, for example, the Y signal and C for one image can be obtained.
When the r signal and the Cb signal are input, since the Y signal is a luminance signal, the code amount of the component is larger than a predetermined value, and the code converted in step ST34 is output. Subsequently, since the second component and the third component are color signals, when the image is a black and white image, it is determined to be equal to the predetermined value,
The converted signal is not output. That is, in the case of a color image, since the Y, Cr, and Cb signals have different independent values, the code amount becomes larger than a predetermined value, and each component is sequentially output. On the other hand, when the color is mono-color, the code amount of the luminance signal (Y signal) is larger than a predetermined value, but the Cr signal and the Cb signal which are color signals are constant, so that they are constant values even when encoded. By determining this constant value, it is possible to determine whether the image is a color image or a monochrome image.

FIG. 5 is a block diagram of the second embodiment of the present invention. The encoding control unit 21 outputs the demultiplexer (D) to the image data distribution range detection unit 24 prior to encoding each component of the input signal held in the image buffer 22.
MPX) 23 is controlled. By this control, the image data stored in the image buffer 22 via the demultiplexer is added to the distribution range detection unit 24. Distribution range detection unit 24
Then, the maximum value and the minimum value of the pixel level of each component are detected.
This detection is performed for each component. Then, the result is input to the significance determination unit 25. The significance determination unit 25 compares the maximum value and the minimum value that are input and determines from the result whether there is significance.

FIG. 6 is a block diagram of the significance judgment unit 25.
The significance determination unit 25 includes a maximum value holding unit 251, a minimum value holding unit 252, and a comparison unit 253. Distribution range detection unit 24
The maximum value detected in 1 is added to the maximum value holding unit 251, and the minimum value is added to the minimum value holding unit 252. Maximum value holding unit 25
1 and each value stored in the minimum value holding unit 252 is the comparison unit 2
Enter in 53. For example, if it is an achromatic image,
The maximum value and the minimum value from the distribution range detector 24 are different for the Y signal, and this value is a constant value of 128 for the Cr and Cb signals. The comparison unit 253 in the significance determination unit 25 detects the coincidences and outputs the result. When the color is not achromatic (does not match), that is, when it is color, 1 is output, and when it is achromatic (match), 0 is output (S
IG signal).

When the SIG signal is 0, the encoding control unit 21 does not instruct the image data encoding unit 26 to perform encoding, but instructs the demultiplexer to determine the significance of the next component. Instruct to select and output the next component again. Further, when the value is 1, it has the significance, and therefore the image data encoding unit 26 is instructed to encode and output the data. By the above operation, the images of all the components are selectively encoded and output.

In the second embodiment of the present invention described above, it is determined whether or not the maximum value and the minimum value are the same for one component, but the present invention is not limited to this.
FIG. 7 is another configuration diagram of the significance determination unit 25. In the above-described embodiment, it is determined whether or not the maximum value and the minimum value match, but in the other configuration diagram in FIG. 7, the range of the maximum value and the minimum value is limited and the result is obtained. .. For example, even a monochrome image is not captured by a monochrome scanner, but when captured by a color scanner, the color signals Cr and Cb of the Y, Cr, and Cb signals naturally have the same value due to an error. Not necessarily. For this reason,
The maximum value and the minimum value obtained by the distribution range detecting unit 24 are stored in the maximum value holding unit 251 and the minimum value holding unit 252. The result is output to the subtraction units 256 and 257. The central part holding part 254 outputs the central value of the values to be invalidated, and the subtracting part 25
6 and 257 obtain the difference between the value and the maximum value and the minimum value stored in the maximum value holding unit 251 and the minimum value holding unit 252.
The obtained results B and C are input to the comparison units 258 and 259.
On the other hand, the threshold value holding unit 255 has respective threshold values, and outputs 1 when the maximum value is greater than or equal to the specific value A than the central value or the minimum value is greater than or equal to the specific value A. Then, the signal is added to the OR gate 260 and is output as a SIG signal obtained by OR addition. If the result of this circuit is within the range of ± A of the value output by the central value holding unit 254, the SIG signal is set to 1 (though the subtraction units 256 and 257 find the difference value). , May have a code). The point is that the SIG signal is set to 0 if the maximum value and the minimum value are within the specific range, and is set to 1 if they are not within the specific range. By performing the significance determination as described above, it is possible to determine the significance even in an image obtained by reading a black-and-white document in which color unevenness is seen, and even if the center value is not 128, there is some deviation. Even if it occurs, it is possible to transmit at high speed and high quality.

FIG. 8 is an operation flowchart of the second embodiment of the present invention. Set I = 1 in step ST41,
In step ST42, the distribution range of the I-th component is detected. The distribution range detection unit 24 detects the distribution range in step ST42.
Processing. Subsequently, in step ST43, the significance of the I-th component is determined. This determination of significance is, for example, the processing of the significance determination unit shown in FIG. 7. Then, it is determined in step ST43 whether or not the result is significant. When it is not significant, that is, when the SIG signal is 0, I to be changed to the next component is changed to I + 1, and the above-mentioned processing step ST is performed.
Execute from 42.

On the other hand, when it is determined in step ST44 that there is a significance (YES), step ST4
5, the I component is encoded, and then the encoded code is S
Output at T46. That is, the image data encoding unit 26 outputs by performing steps ST45 and ST46. The coding control unit 21 subsequently determines in step ST47 whether all the components have ended, and when all the components have not ended, the encoding control unit 21 instructs the next component again, so in step ST48
= I + 1, and the process is repeated from step ST42. When all the components are finished (YES), the whole process is finished.

Even if a specific distribution range is obtained by the above processing, it is determined as a black and white image, so that color unevenness does not occur in the background color and the like, and further effective compression is possible.

FIG. 9 is a block diagram of the third embodiment of the present invention. In the second embodiment of the present invention shown in FIG. 5, the input signals are Y, Cr and Cb signals. On the other hand, in the third embodiment of the present invention, the RGB signals are color-converted and the distribution thereof is obtained.

The RGB signals are input to the color conversion section 31. The color conversion unit 31 converts the RGB signal into a Y signal of a luminance component and Cr, Cb signals of two color components. If the respective values of the RGB signals are R, G, B and the respective values of the Y, Cr, Cb signals are Y, Cr, Cb, the color conversion unit 31.
Is Y = 0.299 × R + 0.587 × G + 0.114 × B Cr = 0.713 × (0.701 × R−0.587 × G−0.114 × B) +128 Cb = 0.564 × (−0.299 × R−0.587 × G + 0.886 × B) +128 Calculate.

These Y, Cr, and Cb signals are applied to the brightness conversion unit 31 on a dot-by-dot basis. The converted results are input to the image buffer 32, and the distribution range detection unit 33 is also provided.
To enter. The distribution range detection unit 33 uses the Y signal and Cr,
This is a detection unit for obtaining a total of three maximum values and minimum values for the Cb signal, and for example, three sets of the configurations of FIGS. 6 and 7 are provided. Then, when one image is input, Y and C respectively
The three maximum values and the three minimum values of the b and Cr signals are added, and the significance determination unit 34 determines whether each value is within the range. Three
Output to 5.

In the third embodiment, the image buffer 3
When inputting to 2 and converting and outputting as a code, for example, when it is determined that the color is achromatic, that is, when the Cr and Cb signals have no significance, the image data encoding unit 3
ADCT conversion is not performed in 6 and only the Y signal is ADCT converted.
Convert and output.

In the above operation, the RG of the input image data
Since the significance is judged in parallel during the conversion of the B signal into the Y, Cr and Cb signals, only the Y signal is coded in the image data coding unit 36, which not only compresses the image but also speeds it up. Processing becomes possible.

[0038]

As described above, according to the present invention, it is possible to suppress code generation of components that do not contribute to a restored image, such as color components in a monochrome image, and to reduce the code amount without degrading the image quality of the restored image. It can be reduced, and the process is completely automated and can be performed without human intervention.

[Brief description of drawings]

FIG. 1 is a processing flowchart of the present invention.

FIG. 2 is a configuration diagram of a first embodiment of the present invention.

FIG. 3 is a detailed configuration diagram of an image data encoding unit.

FIG. 4 is an operation flowchart of the first embodiment of the present invention.

FIG. 5 is a configuration diagram of a second embodiment of the present invention.

FIG. 6 is a configuration diagram of a significance determination unit.

FIG. 7 is another configuration diagram of a significance determination unit.

FIG. 8 is an operation flowchart of the second embodiment of the present invention.

FIG. 9 is a configuration diagram of a third embodiment of the present invention.

FIG. 10 is a conventional configuration diagram.

FIG. 11 is a conventional processing flowchart.

[Explanation of symbols]

 11 image data coding unit 12 coding control unit 13 code buffer 14 code amount calculation unit 15 code output unit 16 code output control unit 21 coding control unit 22 image buffer 23 DMPX 24 distribution range detection unit 25 significance determination unit 26 image Data coding unit 31 Color conversion unit 32 Image buffer 33 Distribution range detection unit 34 Significance determination unit 35 Coding control unit 36 Image data coding unit

Claims (5)

[Claims] 1. A method of encoding the plurality of gradation values in each block obtained by dividing an original image composed of pixels of a plurality of components into a plurality of blocks composed of a plurality of pixels. In determining a significance of each of the plurality of components of the pixel,
And the second step of encoding and outputting only the component determined to be significant in the first step. 2. A method of encoding the plurality of gradation values in each block for each block obtained by dividing an original image composed of pixels of a plurality of components into a plurality of blocks composed of a plurality of pixels. In, there is a first step of encoding all the plurality of components to be input, and a second step of removing the code of an unnecessary component based on the code amount encoded in the first step. An image data encoding method characterized by the above. 3. A method of encoding the plurality of gradation values in each block obtained by dividing an original image composed of pixels of a plurality of components into a plurality of blocks composed of a plurality of pixels. In the first step of detecting the distribution range of each component of the plurality of components to be encoded, and determining whether each of the components is significant from the detection result obtained in the first step. And a third step of encoding only the components determined to be significant in the second step, the image data encoding method. 4. The input original image is an RGB signal, and the first step is to convert the RGB signal of the original image to Y,
4. The image data encoding method according to claim 3, wherein the image data encoding method is performed in parallel when converting into Cr and Cb signals. 5. The determination as to whether or not the second step is significant,
The image data encoding method according to claim 4, wherein it is determined to be significant when the Cr and Cb signals are out of a predetermined range.