JP3403921B2 - Color facsimile machine - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a color facsimile apparatus having a function of receiving a color image.

[0002]

2. Description of the Related Art JPEG is used for color still images having continuous gradation.
(Joint Photographic coding Experts Group: Color still image international standard encoding method)
Coding efficiency is improved by using a color space represented by luminance and color difference components.

Although several such color spaces have been devised, when JPEG is used in a color facsimile, the color reproducibility of an image to be transmitted is impaired due to the color characteristics of an image input / output device such as a scanner or printer. The uniform color space that does not depend on the image input / output device is used.
The international standard color facsimile is required to use the uniform color space CIELAB.

However, in general, the image input / output device uses RGB signals.
It uses a device-dependent color space such as YMC or YMC.
Therefore, in the color facsimile apparatus, it is necessary to convert the color space of the image so as to be compatible with the international standard color facsimile data. Also, especially when outputting images,
Since the color distribution range of the image may exceed the color reproduction range of the image output device, the color space may be converted and color compression may be performed at the same time as the color reproduction range of the image output device. Hereinafter, the color space conversion and the color compression will be collectively referred to as color conversion.

Various methods have been devised for color conversion, not limited to color facsimile machines. However, a natural image scanned by a scanner such as a photograph or painting and an artificial image such as computer graphics have different image characteristics such as a distribution state in a uniform color space. The optimum result cannot be obtained for the image.

As an example of a color conversion method, one using a matrix operation or a lookup table is known. In the color conversion method by matrix calculation, the calculation is basically performed by the following formula (1). It should be noted that the coefficient matrix [D] in the equation (1) is composed of coefficients that minimize the conversion error.

[0007]

[Equation 1] Further, as a method for improving the accuracy, there is a method of performing the calculation by the following equation (2) which is obtained by obtaining the equation (1) up to a high-dimensional term.

[0008]

[Equation 2] In such a matrix calculation, in order to reduce the color error of the entire color space, there is a possibility that the color may be greatly different depending on the place on the color space. In the image mainly composed of characters, the color difference is not so noticeable even if the color conversion is performed by the formula (1). However, natural images such as photographs are not suitable because the difference in color is noticeable depending on the hue. Further, although the calculation with the 3 × 3 matrix can be processed at high speed, the calculation amount increases and the processing becomes slower when the number of terms is increased. That is, the calculation amount of the formula (2) is much larger than that of the formula (1), and thus the time required for the method using the formula (2) is significantly increased as compared with the method using the formula (1). .

The method using a lookup table is as follows:
The conversion source color space is divided into lattices by arranging points forming a cube or a triangular prism, and the correspondence with the corresponding points in the conversion destination color space is used as a lookup table. When the color of the conversion source is other than the grid point, the coordinates in the color space after the conversion are obtained from the ratio of the distances from the neighboring grid points.

According to the method using the lookup table, the conversion error is smaller than that in the matrix operation. However, when the method using the lookup table is implemented by software, the amount of calculation becomes larger than that in the matrix calculation, and the processing time becomes long.

As described above, various existing color conversion methods are
The smaller the conversion error, the larger the amount of calculation and the longer the processing time. For this reason, in a color facsimile apparatus that handles images of various image quality, if a color conversion method with a small conversion error is applied to support high-quality images,
It takes an unnecessarily long time for the color conversion process on the low-quality image. Further, when a color conversion method with a small amount of calculation is applied to give priority to the processing speed, the image quality of a high quality image is deteriorated.

To solve this problem, Japanese Patent Laid-Open No. 8-256261 discloses a technique for determining the type of an image and providing a mechanism for selecting a color conversion method according to the determination so that optimum color conversion can be performed. .

FIG. 8 is a block diagram conceptually showing the structure for realizing the above technique disclosed in Japanese Patent Laid-Open No. 8-256261.

The image data input from the image input unit 21 is given to the image determination unit 22 and the selector 23, respectively. The image determination unit 22 determines whether the image data indicates a character image or a photographic image, for example, from edge detection or density distribution of each color plane. Based on the determination result of the image determination unit 22, the color conversion method selection unit 24 sends the image data showing the character image to the character image color conversion unit 25 and the image data showing the photographic image to the photographic image color conversion unit 26. The selector 23 is switched and controlled so as to give each.

When the image data representing a character image is given from the selector 23, the character image color conversion section 25 performs color conversion of this image data by a method suitable for the character image. Further, when the image data indicating the photographic image is given from the selector 23, the photographic image color conversion unit 26 performs color conversion of this image data by a method suitable for the photographic image. Then, the image output unit 27 outputs the image after the color conversion by the character image color conversion unit 25 or the photographic image color conversion unit 26.

Thus, in the color conversion in the character image color conversion section 25, only the luminance information is converted such as gamma correction, and the saturation information is cut off to speed up the operation. The conversion unit 26 can suppress deterioration of image quality by performing advanced processing including conversion of saturation information.

[0017]

However, in the case where the color conversion processing is performed by receiving the compression-encoded image data as in the color facsimile apparatus, the image determination unit 22 performs the image input in order to perform the image determination. An image decoding unit needs to be provided immediately after the unit 21 to decode the image data. For this reason, a storage device for temporarily storing the decoded image data is required.

However, the amount of data after decoding is 10 megabytes or more in an A4 size image which is a standard size using JPEG used in color facsimiles.
Therefore, in order to store such data, it is necessary to provide a large-capacity memory, which causes an increase in manufacturing cost.

It is desirable for the facsimile apparatus to sequentially decode the received data, perform color conversion, and output the data, from the viewpoint of processing speed and cost, but this is realized by the above method. I can't.

The present invention has been made in consideration of such circumstances, and an object of the present invention is to perform color conversion processing with less error on image data requiring high image quality and image data with low image quality. Decoding and color conversion are received by determining whether the image data to be processed is required to have high image quality while performing both high-speed color conversion processing before decoding the image data. It is an object of the present invention to provide a color facsimile apparatus capable of sequentially performing recorded data.

[0021]

In order to achieve the above-mentioned object, the present invention adds a predetermined parameter class used for decoding the encoded color image data before the encoded color image data. Image decoding means such as an image decoding unit for receiving the facsimile data formed by the above and decoding the encoded color image data using the parameters, and a quantization table or a uniform color space included in the facsimile data. An image quality determination unit such as an image quality determination unit that determines whether the color image data corresponds to a high image quality image with priority on image quality or a low image quality image with priority on processing speed from parameters such as range. If the determination result of the image quality determination means is a high quality image, the high quality image color conversion method is used. In this case, each of the low image quality color conversion methods is selected, for example, the color conversion method selection unit including a color conversion method selection unit and a selector, and the high image quality color conversion method and the low image quality color conversion method. However, it has a function of performing color conversion processing, and converts the color space of the color image data decoded by the image decoding means by the color conversion method selected by the color conversion method selection means, for example, a high-quality color conversion unit and And a color conversion unit including a low image quality color conversion unit.

By taking such means, the high-quality image and the process of the high-quality image of the color image data are processed based on the predetermined parameters added before the color image data encoded by the facsimile data. It is determined in advance which one of the low-quality images prioritizing speed corresponds to, and when this determination result is a high-quality image, the high-quality color conversion method is used, and the determination result of the image quality determination means is low-quality. In the case of an image, the low image quality color conversion method is selected. Then, the color image data decoded from the encoded color image data is subjected to color conversion by the color conversion method selected as described above.

Therefore, before the decoding of the color image data is started, an appropriate color conversion method is selected according to the content of the color image data.

[0024]

DETAILED DESCRIPTION OF THE INVENTION An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing a main configuration of a color facsimile apparatus according to this embodiment. As shown in this figure, the color facsimile apparatus of this embodiment includes a data receiving unit 1, an image decoding unit 2, an image quality determining unit 3, a color conversion method selecting unit 4, a selector 5, a high image quality color converting unit 6, and a low image quality color. It has a conversion unit 7 and an image output unit 8.

The data receiving section 1 demodulates the facsimile transmission signal coming through the communication line C to reproduce the facsimile data, and supplies the facsimile data to the image decoding section 2 and the image quality judging section 13, respectively.

The image decoding section 2 decodes the color image data from the encoded data included in the facsimile data and supplies it to the selector 5.

The image quality determination unit 3 reads the quantization table of the parameters added to the encoded data in the facsimile data, and based on this, the image of the image indicated by the color image data included in the facsimile data. Determine the type. Here, the image quality determination unit 3 distinguishes the image type into a high image quality type that requires highly accurate color conversion and a low image quality type that does not require highly accurate color conversion, and the determination result is the color conversion method. Notify the selection unit 4.

The color conversion method selection unit 4 supplies the color image data of high image quality type to the high image quality color conversion unit 6 based on the determination result of the image quality determination unit 3 and the color image data of low image quality type. Is controlled to switch the selector 5 so as to apply to the low image quality color conversion unit 7.

The selector 5 selects either the high image quality color conversion unit 6 or the low image quality color conversion unit 7 under the control of the color conversion method selection unit 4 and performs image decoding on the selected color conversion unit. The color image data given from the section 2 is given.

The high image quality color conversion section 6 performs color conversion processing by a color conversion method with few conversion errors on the color image data given from the selector 5. Examples of the color conversion method used by the high image quality color conversion unit 6 include a method using a look-up table and a method using a high-order matrix such as the expression (2). Then, the high-quality color conversion unit 6 outputs the color image data after the color conversion processing to the image output unit 8
Give to.

The high image quality color conversion section 7 performs color conversion processing by a color conversion method which requires less calculation amount than the high image quality color conversion section 6 on the color image data given from the selector 5. As a color conversion method used by the low image quality color conversion unit 7, for example, there is a method using a 3 × 3 matrix as shown in the equation (1). Then, the high image quality color conversion section 7 gives the color image data after the color conversion processing to the image output section 8.

The image output unit 8 is formed by using, for example, a color printer or the like, and an image represented by color image data provided from either the high image quality color conversion unit 6 or the low image quality color conversion unit 7 is recorded on a recording paper. Record and output.

Next, the operation of the color facsimile apparatus configured as described above will be described.

First, in the facsimile data of the color facsimile, a large number of parameters for decoding the color image data from the encoded data are added before the encoded data obtained by encoding the color image data for data compression. To be done.

FIG. 2 is a diagram showing a part of the data format of facsimile data. As shown in this figure, each parameter is marked with a marker for each type. Then, some of these parameters can be omitted, and in this case, a predetermined default value is used. The other parameters are indispensable, but there are preset standard values for these as well. For each parameter, it is possible to use the default value and standard value as they are, but it is also possible to change the value according to the image to be transmitted.

The quantization table, which is one of the parameters, is determined when transmitting facsimile data. Although the contents of the quantization table are not defined by the color facsimile standard, the JPEG standard normally recommends the recommended values as shown in FIG. And by changing the value in the quantization table,
It is possible to control the size of encoded data. That is, the quantization step shown in the quantization table is a parameter for removing high frequency components in a block to reduce the amount of information at the time of JPEG encoding. The larger the step width, the higher the compression ratio, but the larger the error between the decoded color image data and the original color image data.

As for the value of the quantization step described in the quantization table, in the normal usage, the recommended value is used as it is, or the product of each quantization step and the image quality coefficient Q which is a positive real number is used. To use. The smaller the value of the image quality coefficient Q, the narrower the quantization step width, and the image quality is improved. When the image quality coefficient Q is 0.5 or less, no difference from the original image can be seen. The quantization step of facsimile data is often the product of the recommended value and the image quality coefficient Q.

Therefore, on the side of transmitting the facsimile data, when the transmission is performed with the image quality given priority, the operation of reducing the value of the quantization step described in the quantization table is performed. Further, in the case of transmission with speed priority, a large quantization step is used to reduce the amount of data after encoding, and the communication time is shortened.

Therefore, the image quality judging section 3 is, as shown in FIG.
The quantization table included in the received facsimile data is read to check the quantization step width (step ST
1). At this time, only the quantization table of the luminance component that greatly affects the image quality may be the target of the determination, or both the quantization tables of the luminance component and the color difference component may be the targets.

Subsequently, the image quality judgment unit 3 compares the quantization table read in step ST1 with a predetermined reference quantization table (step ST2). As the reference quantization table, the one shown in FIG. 3 recommended by JPEG may be used, or may be uniquely determined.

Here, the comparison between the quantization table read in step ST1 and the reference quantization table is specifically performed as follows. That is, the quantization table has 6
Since four step widths are included, step ST1
With the i-th quantization step width of the quantization table read in step 1 as Ti and the i-th quantization step of the reference quantization table as TOi, the following equation (3) is calculated.

[0043]

[Equation 3] If the calculation result A of the equation (3) is a positive value, it is determined that the quantization table read in step ST1 is larger than the reference quantization table, and if the calculation result A is a negative value, it is determined in step ST1. It is determined that the read quantization table is smaller than the reference quantization table.

If the quantization table read in step ST1 is larger than the reference quantization table, the image quality determination section 3 determines that it is of high image quality type (step ST).
3). If the quantization table read in step ST1 is smaller than the reference quantization table, the image quality determination unit 3 determines that the image quality type is low (step ST4).

The color conversion method selection unit 4 controls the switching of the selector 4 based on the result of the determination made by the image quality determination unit 3 as described above.

After this, the coded data is given to the image decoding unit 2, and the image decoding unit 2 decodes the color image data. As a result, the color image data obtained by the image decoding unit 2 is converted into the high image quality color conversion unit 6 and the low image quality color conversion unit 7.
It is given to the one selected by the selector 5 as described above. Then, the color image data is subjected to color conversion processing by the high image quality color conversion unit 6 or the low image quality color conversion unit 7, and then provided to the image output unit 8 to be used for image output.

Thus, according to the present embodiment, for the color image data of the image quality priority transmitted with the quantization step set to be small in order to reduce the error due to the encoding / decoding, the high image quality color conversion unit 6 is performed. Highly accurate color conversion is performed by a color conversion method with a small conversion error. Further, for the color image data of speed priority transmitted with a large quantization step set in order to shorten the communication time, the low image quality color conversion unit 7 can perform high-speed color conversion by a color conversion method with a small calculation amount. Done. As a result, it is possible to achieve both color conversion processing with less error for image data that requires high image quality and high-speed color conversion processing for image data that has low image quality.

In the present embodiment, the decoding is performed before the image decoding unit 2 decodes the color image data from the encoded data based on the quantization table of the parameters that arrive prior to the encoded data. A color conversion method suitable for color-converting the obtained color image data is selected. As a result, the received facsimile data can be sequentially decoded, and the color conversion can be performed for each minimum unit of the image block to be decoded and sent to the image output unit 8, and an extra memory is not required.

The present invention is not limited to the above embodiment. For example, in the above-described embodiment, the image quality determination unit 3 determines the image type of the image represented by the color image data based on the quantization table, but determines the image type based on another parameter included in the facsimile data. You can also do it.

For example, the range of the uniform color space, which is one of the parameters, can be used. Color image data on the uniform color space is 0 to 25 for encoding processing.
Must be mapped to an integer of 5. The range and offset are used in the mapping calculation.

Equation (4) below is the uniform color space CIELAB.
Is a calculation formula for mapping each component of 0 to 255. Here, L is the value of the L ^* component of CIELAB, NL is the value of the mapped L ^* component, RANGE and OFFSE.
T is the range and offset of the L ^* component, respectively.

[0052]

[Equation 4] The range and offset of the uniform color space are set as default values as shown in FIG. As this default value is set so large that most of the colors existing in the world are included, it is larger than the color reproduction range of the image input / output device actually used in the facsimile, such as a scanner or printer. . Even if the color range of the image to be encoded is narrow, if encoded with the default value, the gradation information will be reduced when converted to the uniform color space, and the number of gradations of the color image data decoded on the receiving side will also decrease. The error from the original image becomes large. When the range of the uniform color space is matched with the color range of the coded image, the coded image can be coded without losing the gradation information, so that a decoded image with less error can be obtained.

Therefore, on the side of transmitting the facsimile data, in the case of transmitting the image with priority to the image quality, the value of the quantization step described in the quantization table is reduced and the uniform color space range is adjusted to the image color range. To do. Further, in the case of transmission with speed priority, a large quantization step is used to reduce the amount of data after encoding, and the communication time is shortened.

Therefore, as shown in FIG. 6, the image quality judgment section 3 reads the range of the uniform color space included in the received facsimile data (step ST11) and compares this with a default value (which may be a recommended value). If all the elements are the same (step ST12), the image quality is determined to be low (step ST13), and if any one is different, the image quality is determined to be high (step ST14).

The image type determination in the image quality determination unit 3 may be performed based on both the quantization table and the uniform color space range. That is, as shown in FIG. 7, the image quality determination unit 3 reads the quantization table included in the received facsimile data (step ST21), compares it with the reference quantization table (step ST22), and includes it in the facsimile data. If the quantization table and the reference quantization table are equal, the range of the uniform color space included in the facsimile data is further read (step ST23), and this is compared with the default value (which may be a recommended value) (step ST23). ST2
4). When the quantization table included in the facsimile data is larger than the reference quantization table by the image quality determination unit 3, or all the elements are the same in the uniform color space range and the default value included in the facsimile data. In this case, the image quality is judged to be low (step ST
25) If the quantization table included in the facsimile data is smaller than the reference quantization table, or if any one of the elements of the uniform color space range and the default value included in the facsimile data is different. It is determined that the image quality is low (step ST26).

Besides, various modifications can be made without departing from the scope of the present invention.

[0057]

According to the present invention, facsimile data obtained by adding predetermined parameters used for decoding the coded color image data before the coded color image data is received and the parameters are received. Image decoding means for decoding the coded color image data using a class, and a high-quality image in which the color image data has high image quality priority and a low-quality image in which processing speed has priority based on the parameters included in the facsimile data. Image quality determining means for determining whether or not the image quality determining means determines that the image quality determining means determines that the image quality determining means is a high quality image , A color conversion method selecting means for selecting a low image quality color conversion method, the high image quality color conversion method, and the low image quality color conversion method Any of them has a function of performing color conversion processing, and is provided with color conversion means for converting the color space of the color image data decoded by the image decoding means by the color conversion method selected by the color conversion method selection means. Therefore, the image data to be processed is required to have high image quality while achieving both the color conversion process with less error for the image data requiring high image quality and the high-speed color conversion process for the image data having low image quality. By determining whether or not there is the image data before decoding the image data, the color facsimile apparatus can sequentially perform decoding and color conversion on the received data.

[Brief description of drawings]

FIG. 1 is a block diagram showing a main configuration of a color facsimile apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing a part of a data format of facsimile data.

FIG. 3 is a diagram showing the contents of a quantization table defined as recommended values in JPEG.

FIG. 4 is a flowchart showing a processing procedure relating to image type determination by an image quality determination unit 3 in FIG.

FIG. 5 is a diagram showing default values of a range and an offset in a uniform color space.

FIG. 6 is a flowchart showing a first modified example of a processing procedure relating to image type determination by an image quality determination unit 3 in FIG.

7 is a flowchart showing a second modified example of the processing procedure relating to the image type determination in the image quality determination unit 3 in FIG.

FIG. 8 is a block diagram conceptually showing a configuration for realizing the above technique disclosed in Japanese Patent Laid-Open No. 8-256261.

[Explanation of symbols]

1 ... Data receiving unit 2 ... Image decoding unit 3 ... Image quality judgment unit 4 ... Color conversion method selection section 5 ... Selector 6 ... High quality color converter 7 ... Low image quality color converter 8 ... Image output section

─────────────────────────────────────────────────── ─── Continuation of the front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) H04N 1/46 G06T 1/00 510 H04N 1/41 H04N 1/60

Claims (4)

(57) [Claims] 1. Facsimile data obtained by adding predetermined parameters used for decoding the coded color image data before the coded color image data and receiving the code by using the parameters. Image decoding means for decoding the converted color image data; and the color image data corresponding to either a high-quality image with priority on image quality or a low-quality image with priority on processing speed from parameters included in the facsimile data. Whether the image quality determination means determines whether the image quality determination means is a high quality image, or the image quality determination means determines whether the image quality determination means is a low quality image. Color conversion method selecting means for selecting each of the image quality color conversion methods, and color conversion using any of the high image quality color conversion method and the low image quality color conversion method. And a color conversion unit for converting the color space of the color image data decoded by the image decoding unit by the color conversion method selected by the color conversion method selection unit. Color facsimile machine. 2. The image quality determination means compares a reception quantization table included in the parameters with a reference quantization table which is predetermined as a reference value, and the reception quantization is performed from the reference quantization table. 2. The image is determined to be a low quality image if the quantization step in the table is larger, and is determined to be a high quality image if the quantization step in the reception quantization table is smaller than the reference quantization table. The color facsimile machine according to 1. 3. The image quality determination means includes received color space range information included in parameters and indicating a range of a uniform color space,
Compared with the reference color space range information predetermined as a reference value, if the reference color space range information and the reception color space range information are the same, it is determined as a low image quality image,
The color facsimile apparatus according to claim 1, wherein if the reference color space range information and the received color space range information are different, it is determined that the image is a high quality image. 4. The image quality determination means compares a reception quantization table included in the parameters with a reference quantization table which is predetermined as a reference value, and the reception quantization is performed from the reference quantization table. If the quantization step in the table is larger, it is determined to be a low-quality image, and if the quantization step in the reception quantization table is smaller than the reference quantization table, it is determined to be a high-quality image, and the reference quantization table is used. If the quantization step is the same in the reception quantization table, the reception color space range information indicating the range of the uniform color space included in the parameters and the reference color space previously determined as the reference value The range information is compared, and if the reference color space range information and the received color space range information are the same, it is determined that the image is of low image quality. The color facsimile apparatus according to claim 1, wherein if the reference color space range information and the received color space range information are different, it is determined that the image is a high quality image.