JPH0435361A - Picture coding device and picture coding and decoding device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a single image encoding/decoding device, and particularly to an image encoding/decoding device that handles images including text areas and photographic areas.

[Conventional technology] Conventional techniques for efficiently compressing image signals include 2.
Many encoding/decoding processing techniques suitable for value images, multi-value images, color images, etc. are known.

For example, for binary images, MH
(Modified Huffman), MR (Modified Read), etc. are known.

Furthermore, for multivalued images, a technique using orthogonal transformation processing is known, and by applying this to three colors, a color image can be encoded and decoded.

These methods are described, for example, in 198 of the Journal of the Institute of Electronics, Information and Communication Engineers.
It is explained from page 657 of July 1988 issue (vol. 71. No. 7).

However, a single image often includes image regions with different properties such as binary images and multivalued images, and using only one type of these encoding/decoding processing methods will result in compression problems. Not enough effect.

On the other hand, depending on the nature of the image. As a region separation method, for example, a technique described in Japanese Patent Laid-Open No. 59-55581 is known.

This technology uses the results of two-dimensional Fourier transformation of an image to
This method realizes region segmentation using the position and absolute value of the signal peak point.

[Problems to be Solved by the Invention] However, when the above-mentioned conventional technology is applied as is to an image encoding/decoding device, it is necessary to use means for performing Fourier transform of the image signal in order to segment image regions according to their properties. In particular, it is necessary to provide such a device, which causes problems such as increased cost and increased size of the device.

Furthermore, the method of dividing regions using the peak points of the results of two-dimensional Fourier transformation utilizes the regularity of rows and columns in a document, and does not utilize the properties of binary images in general. This method is not necessarily optimal for dividing areas between general binary images and multivalued images.

SUMMARY OF THE INVENTION Therefore, the present invention provides an image code that can perform encoding and decoding processing according to the properties of binary images and multivalued image areas, respectively, without significantly increasing the size of the device. The purpose is to provide an encoding/decoding device.

[Means for Solving the Problems] In order to achieve the above object, the present invention includes a discrete cosine transform means and an input image that is separated into a binary image region and a multi-value image region based on the transformation result of the discrete cosine transform means. An image separation unit having a separation unit, a binary image encoding unit that encodes an image of the binary image area separated by the image separation unit, and the discrete cosine transformation unit shared with the image separation unit, Provided is a multilevel image encoding unit that encodes an image of a multilevel image area separated by a separation unit; and an image encoding device that encodes.

Note that in this image encoding device, the separating means includes:
After determining the type of image block consisting of about 8 x 8 pixels as a binary image or a multivalued image, it is desirable to perform an integration process on consecutive image blocks of the same type, and then perform separation on the integrated area. .

Further, the encoded data of the image of each region is
It is desirable to manage the image along with area management data including type information indicating that the image is a value image or a multi-value image and spatial information of the area.

Further, when separating the color image into the regions, the separation means converts the input image into two parts based on the conversion result of the discrete cosine conversion means of one color signal or luminance signal among the color signals constituting the color image. The image may be separated into a value image area and a multivalue image area.

Furthermore, the multi-valued image encoding unit is configured to perform a binary image encoding unit on each of a binary image area and a multi-valued image area specified from the outside instead of the area separated by the image separation unit. An image in the binary image area.

The multi-value image encoding unit may encode the image in the multi-value image area.

Further, it is preferable that the multivalued image encoding section includes code word conversion means for converting the conversion result of the discrete cosine transform means into code words.

Further, it is preferable that the binary image encoding unit has code word conversion means for converting the code word of the image in the binary image area based on run length measurement in the scanning line direction.

In addition, when a single image memory is provided, it is desirable to share the memory used for binary image encoding processing and the memory device used for multilevel image encoding and processing.

It also has a data input/output means for communicating with a host device such as a processor, and can receive commands from the processor to initialize the internal state, start encoding or decoding processing, interrupt processing that is currently being executed, and target the encoding processing. Setting of the image area, etc. may also be accepted.

Furthermore, as the image data input/output means to the image encoding/decoding device, it is desirable to share data input/output means for binary images and multivalued images.In this case, data representing one pixel of the multivalued image Data representing multiple pixels of a binary image may be assigned to the same bit string.

Although it is preferable to use a discrete cosine transform means, in some cases, other orthogonal transform means may be provided in place of the discrete cosine transform means.

Further, it is desirable that this image encoding device be built into a semiconductor integrated circuit.

In order to achieve the above-mentioned object, the present invention also provides a binary image decoding unit that decodes an image in an encoded binary image area and an image in a multi-value image area encoded by discrete cosine transform. The present invention provides an image decoding device including a multivalued image decoding unit for decoding, and an image encoding/decoding device for encoding the image encoding device or the semiconductor integrated circuit.

Note that in this image encoding/decoding apparatus, the image decoding apparatus may include a converting means for converting a multivalued image into a binary image.

Further, it is desirable that the image encoding/decoding device or the image encoding device include means for enlarging or reducing an image.

The present invention also provides a communication control unit that controls the transmission of encoded image data encoded by the image encoding device or the semiconductor integrated circuit, and the image encoding device or the semiconductor integrated circuit provided therein. Provided is an image communication device that encodes information that the user has. In addition.

This image communication device may also include the decoding device, if necessary.

Furthermore, the present invention provides the image encoding device that encodes an input image or the semiconductor integrated circuit, and a memory that stores encoded image data encoded by the image encoding device or the semiconductor integrated circuit provided therein. An image storage device is provided, comprising: means. Note that this image storage device may also include the decoding device as required.

[Operation] The operation will be explained below by taking a typical image encoding device and image encoding/complexing device according to the present invention as an example.

According to the image encoding device according to the present invention, from the transformation result of the discrete cosine transformation means shared with the multilevel image encoding section,
An input image is separated into a binary image area and a multivalued image area.

Therefore, there is no need to provide any means for image region separation.

Further, after determining the type of image block consisting of approximately 8 x 8 pixels as a binary image or a multivalued image, continuous image blocks of the same type are integrated, and the integrated area is separated. By doing so, the discrete cosine transform means can be configured easily and compactly. Further, each image area can be encoded according to its properties.

Further, the encoded data of the image of each region is
If it is managed together with area management data including type information indicating that it is a value image or multi-value image and spatial information of the area,
Image layout processing etc. can be easily performed.

In addition, when separating a color image into the regions, the input image can be divided into binary image regions and multiple regions based on the conversion result of the discrete cosine transform means of one color signal or luminance signal among the color signals constituting the color image. By separating the image into value image regions, the image region separation process can be simplified.

Further, the multi-valued image encoding unit receives two
For each of the value image area and the multi-value image area, 2
The flexibility of the apparatus can be increased by adding a function of encoding images in a binary image region in the value image encoding section and encoding images in the multivalue image region in the multivalue image encoding section.

Furthermore, according to the image encoding/decoding device according to the present invention,
The binary image decoding section decodes the encoded image of the binary image region, and the multi-value image decoding section decodes the image of the multi-value image region encoded by discrete cosine transform. Also,
In this image encoding/decoding device, if the image decoding device is equipped with a conversion means for converting a multivalued image into a binary image, the image encoding device can be used for printing devices that can only print binary images, for example. can be output.

Furthermore, the image communication device and the image storage device according to the present invention can transmit or store image data that is efficiently encoded according to the properties of the image for each area, so that the communication fee, time, or storage capacity is reduced. etc. can be reduced.

(The following is a margin) [Embodiment 1 Below, an image encoding device and a converted image encoding device according to the present invention will be described.
An embodiment of the decoding device will be described using a single image encoding/decoding device as an example.

First, a coding method used in the image encoding/decoding apparatus according to this embodiment and a method for dividing regions into binary images and multivalued images will be explained.

In this embodiment, a plurality of encoding means for binary image regions and a plurality of encoding means for multivalued image regions are provided. Discrete cosine transform (DCT: Disc
A method is adopted that combines the rete Co51ne Transform) and entropy encoding.

DCT is defined, for example, by the following formula.

If the input signal is x (m) and the output signal is X (k), then k=o, 1. ..., N-1 The output signal X(k) represents a correlation value with a cosine function from direct current to high frequency.

In image encoding, signal processing is performed using the bias in the distribution of frequency components of an image signal.

Incidentally, this signal component after the discrete cosine transform can be used to determine the properties of the image for dividing the image into regions.

That is, the image signal generated by inputting a monochrome binary image has a signal level equivalent to a white background (in most cases, the background density of paper) and black (printing ink density).

On the other hand, an image signal generated by inputting a photographic image has a signal distribution including an intermediate level.

By performing discrete cosine transformation on such an image signal, a large amount of direct current components or high frequency components are generated in a black and white binary image, while fewer high frequency components are generated in a photographic image.

This property can also be obtained by other orthogonal transformations, but it is empirically known that this property can be obtained significantly by discrete cosine transformation. Therefore, in this embodiment, the frequency distribution obtained by this discrete cosine transform is used to perform image region segmentation that is easier and more reliable than the region segmentation using the Fourier transform according to the prior art. .

In this embodiment, the conditions for determining the image properties based on the distribution of the discrete cosine transform results are set by performing measurement experiments on a large number of samples in advance and using the statistical values of the experimental results.

As described above, in this embodiment, the discrete cosine transform is used as the encoding method, and the discrete cosine transform is also used to segment the image regions. Therefore, in the apparatus, the discrete cosine transform means, which is a means for encoding a multivalued image, can also be used for dividing the image region, without specifically providing an orthogonal transform means for dividing the image region. There is no need to significantly increase the cost and scale of the device.

Here, details of region segmentation using the frequency distribution obtained from the discrete cosine transform will be explained.

Discrete cosine transform in encoding processing performs image block unit processing, but from the viewpoint of compression rate.

The shape of the image block is a square of 8×8 pixels.

Therefore, it is desirable to perform region segmentation using the results of the discrete cosine transform on a block-by-block basis.

By the way, when the block is small, for example in a character area, the image signal of the image block of the character line spacing and dose contains only a DC component representing the background density of the paper, and
Since high-frequency components appear in image blocks containing characters, it is difficult to separate character regions with a high degree of organization based on the determination results of only individual image blocks.

Therefore, in this embodiment, a global area consisting of a set of a plurality of image blocks is generated. That is, a region is generated based on the type appearance frequency of a plurality of adjacent image blocks using the type determination result for each image block.

This situation is shown in FIG.

That is, first, in a binary area such as a character area, an image is a collection of pixels that indicate the background density of the paper and the density of printing ink. (1) Everything within the image block is the background density of the paper.

■All image blocks are printed ink density.

■The above and ■ appear in the image block.

It will be one of the following.

Next, in a multivalued image area such as a photograph,
Many intermediate concentrations appear.

Therefore, the energy distribution for each band corresponding to these image blocks is determined in advance through measurement experiments using a large number of images as samples, and is used as the determination condition 304.

Then, the frequency distribution resulting from the discrete cosine transform of each block is divided into predetermined bands 201, the average power for each band is calculated 302, and if it corresponds to the above, it is converted into a binary image. If it is legal and corresponds to ■, it is determined that the multivalued image is legal 303
゜And, as long as a statistically large number of image blocks with binary image or multivalue image legal determination results appear in a certain area, that area is composed of image blocks of the same type. and separate that area.

Note that the size of this area may be set to a fixed size in advance, or it may be set horizontally,
A global image area may be created by increasing the number of vertically adjacent image blocks. This processing procedure. It is shown in Figure 4.

That is, for example, while gradually enlarging the image area from the image block serving as the base point (step 405), the types of image blocks included in the image area are determined to have a legal distribution to be separated as areas of the same type. If it is legal, the image area is further expanded by sequentially connecting image blocks in the horizontal and vertical directions from the reference image block (step 402). ), on the other hand, when an image block of an illegal type begins to appear, the enlargement of the image area may be terminated (step 403), and the image area may be separated according to the type that is legal (step 403). 405, 406).

The configuration of the encoding section of the image encoding/decoding apparatus according to this embodiment will be described below.

FIG. 1 shows this configuration.

In the figure, 101 is an original image buffer, 102 is a discrete cosine transform means, 103 is an intermediate buffer, and 105 is a type determination means 1.
05 and 106 are area segmentation means, 107 is structured data 1
07, 110 is a switching device, 111 is an encoding means that is a combination of MHlMR or MMR encoding means for encoding black and white binary images, multi-value encoding means for encoding multi-value images, etc., and 112 is a structured encoding means. It is a code buffer.

Next, the encoding operation of the image encoding/decoding apparatus according to this embodiment in this configuration will be explained.

FIG. 6 shows the operating procedure.

The original image buffer 101 stores image signals from image input devices such as scanners and TV cameras. It is desirable that the storage capacity is for one screen or more, but it may be for one screen or less.

The discrete cosine transform means 102 uses the image signals accumulated in the original image buffer 101 to execute a discrete cosine transform process for each image block of approximately 8×8 pixels, and then accumulates the result in the intermediate buffer 103.

In this embodiment, the storage capacity of the intermediate buffer 103 is the same size as that of the original image buffer 101.

The type determining means 105 inputs the original image signal from the original image buffer 101 and the discrete cosine transform result from the intermediate buffer 103 in order to determine the image properties of each image block. The distribution of frequency components is determined based on predetermined determination conditions, and the type of the image block is output.

Using the type determination result for each image block obtained in this way, the area separation means 106 determines whether or not the types of adjacent image blocks are legal to be combined as the same area. A global region, that is, a region composed of a plurality of image blocks, is generated and the region information is output (step 601).

Here, as mentioned above, the determination of whether or not it is legal is done by statistically calculating the appearance frequency of image blocks that occur in areas that can be determined to be the same area, using a large number of image samples in advance or based on measurement experiments. By summarizing the
By comparing this statistical value with the processing results of the actually input image information, it is determined that the image blocks are legal if they can be determined to be from the same area, and illegitimate if the image properties can be determined to be different. , generate the above image area.

Note that in this embodiment, the image area is assumed to be rectangular for convenience, but the shape may be arbitrary.

Information such as the identification of the document area or the image area, and the position and size of the area outputted by the area separation means 106 is stored as structured data 107, and an externally installed software program etc. The layout structure is determined and simultaneously transmitted to the switching device 110 (step 602).

In order to perform encoding processing suitable for the nature of the image area, MHlMR or M
The original image buffer 5
The original image signal from 01 is sent (steps 603, 604
), the encoding process is executed according to the type of document and image in the region (step 605), and the resulting encoded data is organized by image region and structured code) < buffer 1
12 (step 606).

On the other hand, application software installed externally outputs the document layout structure from the structured data 107,
By using the output of the structured code buffer 112, the image information of each region of the layout structure of the input image can be efficiently compressed and used for transmission or storage purposes.

Figure 7 shows an overview of the use of this layout structure.

It is shown in FIG.

First, as shown in FIG. 7, the layout structure is analyzed and determined from structured data obtained from an image 701 inputted from a scanner, etc., and a layout structure is created.
02° Then, if necessary, the position and size of each area in the layout structure number is transmitted and stored as header information in combination with the code data of the corresponding area.

In this way, if there is a need for re-editing by handling the layout structure for each image manually using regions as units, each region in the determined layout structure can be changed, moved, or added as a unit.

7 to be able to perform processing such as startup and enlargement.
03゜By the way, in region separation of the input image (1), for example,
There are cases where the same encoding process is performed on the entire screen and specific control such as 1 is required. Therefore, in this embodiment, the control parameters from the external device are transmitted to the type determining means 105. and specify the desired behavior.

As described above, in this embodiment, the encoding means 111
When using a method that combines discrete cosine transform and entropy encoding as the encoding process of the multi-encoding means,
By executing the discrete cosine transform in the discrete cosine transform means 102 and storing the result in the intermediate buffer 103, the contents of the intermediate buffer 103 are transmitted to the multilevel encoding means through the switching means 110 and entropy encoded. By performing only the signal processing equivalent to
Realizes encoding processing of multilevel signals.

That is, conventionally, multivalued image encoding means includes discrete cosine transform means 501 and quantization means 502 as shown in FIG.
, entropy encoding means 503, but in this embodiment, only quantization means and entropy encoding means are provided as multilevel image encoding means, and discrete cosine transformation means 501 performs signal processing for region separation. It is shared with

In addition, in the discrete cosine transform means, by setting the coefficients of the transform processing, it is possible to switch between discrete cosine transform, discrete sine transform, discrete Fourier transform, etc.
Each of these orthogonal transformations can also be used for image block determination.6 Note that the encoding unit of the image encoding/decoding device according to this embodiment can also be configured as shown in FIG. It is. That is, the discrete cosine transform means 10
The embodiment of FIG. The intermediate buffer 103 that was necessary in
This is an example of a configuration in which 04 is installed.

Next, FIG. 9, which explains the decoding section of the image encoding/decoding apparatus according to this embodiment, shows the configuration of this decoding section.

In the figure, 901 is a code data input means for receiving code data, 902 is a plurality of decoding means corresponding to the encoding unit, and 903 is a whole image data that is decoded for each area by the layout structure data and the decoding means. This is an image data reconstruction means for reconstructing an image.

By the way, when decoding code data regarding a multilevel image,
In an image communication device that records reproduced image data,
If only a recording device for recording a binary image is provided, as shown in FIG. 10(a), a means for level converting the multivalued data reproduced using a threshold value is provided, for example,
By performing pseudo halftone processing and performing pseudo halftone recording using a binary recording device, a recorded image similar to a multivalued image can be obtained. Here, systematic dithering, error dispersion dithering, etc. can be used as pseudo halftone processing.

This allows the receiving side to perform binary recording and multilevel recording according to the characteristics of the recording device on the receiving side, without having to encode and transmit image data that has been subjected to pseudo-halftone processing of a multilevel image on the transmitting side. You can select and obtain high quality recorded images.

Note that, according to the image encoding/decoding apparatus according to this embodiment, by representing image data as a multivalued signal, the image data shown in FIG.
), one image area can be easily enlarged or reduced by performing interpolation processing on multivalued signals.

Therefore, editing processing of image areas can be easily realized.6 Also, when targeting a color image, it is possible to process multiple color signals that make up a color signal as independent multi-valued image signals. , which can be easily encoded and decoded. In this case.

Image region separation may not be performed for all of the plurality of color signals, but region separation may be determined for at least one color or luminance signal.

In addition, in the input/output of image data of the image encoding/decoding device according to this embodiment, if a frame memory for storing image data is provided separately for a multi-value image and a binary image, the multi-value image memory It is desirable that the image data and the binary image memory be configured by bus connections as shown in FIG. 11, and that efficient transfer of image data can be realized by switching between them.

The illustrated example shows an example in which each is connected by a 16-bit bus.

In addition, when the encoding means and the decoding means in the image encoding/decoding apparatus according to this embodiment are configured as LSI etc., as shown in FIG.
), by specifying its vertical position, it is possible to process one line as a unit, and for multilevel images, as shown in the same figure (b), by specifying the vertical and horizontal positions of the base pixel, it is possible to process an image block as a unit. A means may be provided. This is because the apparatus can be easily configured without providing any means for generating addresses outside the encoding/decoding means. Note that within the encoding means and decoding means, processing can be switched on a line-by-line basis and on an image-block basis using the region separation result indicating whether the target image region is binary or multi-valued. .

Finally, FIG. 15 shows an image signal communication device and an image storage device using the image encoding/decoding device according to this embodiment.

FIG. 15(a) shows an image signal communication device, and 130 in FIG.
1 is an image input device such as a scanner, 1302 is an image printing or display device such as a printer, 1304 is an image encoding/decoding device according to this embodiment, and 1305 is a transmission control device that controls communication of encoded data. .

Further, FIG. 15(b) shows an image storage device.

In the figure, 1301 is an image input device such as a scanner, 1302 is an image printing or display device such as a printer, 1304 is an image encoding/decoding device according to this embodiment, and 1306 is an optical disk device or the like that stores encoded data. It is an image storage device.

As described above, according to this embodiment, efficient compression can be achieved by dividing an image input by a scanner or the like into regions and performing encoding processing suitable for each region.

Further, by performing the above region separation based on the frequency distribution of each image region obtained by discrete cosine transformation processing, accurate determination using image properties can be performed. Although it is desirable to use a discrete cosine transform process to determine this image property, even if an orthogonal transform process other than this is used, the determination is made based on the frequency distribution, which is not the case with the prior art described above. It is possible to perform more accurate determination than determination using peak values.

Further, by using the signal processing of the discrete cosine transform as the discrete cosine transform used in the encoding process of a multivalued image, it is possible to simplify the signal processing procedure and simplify the apparatus.

Furthermore, there is also the effect that image re-editing, storage, transmission, etc. can be efficiently realized using the determination result of the layout structure of one document image.

Note that when using an orthogonal transform other than the discrete cosine transform, this orthogonal transform is used and also used in the encoding process of the multivalued image. Or, conversely, the orthogonal transformation used in the encoding process of the multivalued image is also used to determine the image properties.

[Effects of the Invention] As described above, according to the present invention, it is possible to encode and encode regions of binary images and multivalued images according to their properties, without significantly increasing the size of the device. An image encoding/decoding device that can perform decoding processing can be provided.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of the encoding section of an image encoding/decoding device according to an embodiment of the present invention, and FIG. 2 shows another configuration of the encoding section of the image encoding/decoding device. 3 is an explanatory diagram showing the type determination process of a pixel block diagram, FIG. 4 is a flowchart showing the processing procedure for creating bizarre images, and FIG. 5 shows the configuration of a conventional multi-level image encoding means. 6 is a flowchart showing the steps of the encoding process of the image encoding device, FIGS. 7 and 8 are explanatory diagrams showing the concept of the layout structure and how to use it, and FIG. 9 is the image encoding A block diagram showing the configuration of the decoding unit of the decoding device. Fig. 10 is an explanatory diagram showing the conversion between multi-valued images and binary images, and image enlargement and reduction processing. Fig. 11 is the image memory encoding/decoding. A block diagram showing the connection with the device, FIG. 12 is an explanatory diagram for explaining the processing unit of encoding and decoding processing of binary images and multivalued images, and FIG. 13 is an image communication device and an image storage device. FIG. 2 is a block diagram showing the configuration of FIG. 401... Original image buffer, 402... Image block input means, 403... Discrete cosine transformation means, 404...
... Quantization means, 405 ... Entropy encoding means, 501 ... Original image buffer, 502 ... Discrete cosine transformation means, 503 ... Intermediate buffer, 504 ...
Switch, 505... Type determination means, 506... Area determination means, 507... Structured data, 510...
Switching means, 511... Encoding means, 512...
Structured code buffer.

Claims (1)

[Scope of Claims] 1. An image separation unit having a discrete cosine transformation means and a separation means for separating an input image into a binary image region and a multivalued image region based on the transformation result of the discrete cosine transformation means; A binary image encoding unit that encodes an image of a binary image area separated by the image separation unit and the discrete cosine transform means shared with the image separation unit are used to convert the image of the multivalued image area separated by the image separation unit. An image encoding device comprising: a multivalued image encoding unit that performs encoding; 2. The image encoding apparatus according to claim 1, wherein the separation means determines the type of image block consisting of about 8×8 pixels as a binary image or a multivalued image, and then separates consecutive images of the same type. An image encoding device characterized by performing an image block integration process and separating the integrated regions. 3. The image encoding device according to claim 1 or 2, wherein the encoded data of the image of each region is combined with type information indicating that the region is a binary image or a multivalued image and the space of the region. What is claimed is: 1. An image encoding device comprising means for managing area management data including information. 4. The image encoding device according to claim 1, 2 or 3, wherein when separating the color image into the regions, the separation means selects one of the color signals constituting the color image or An image encoding device characterized in that an input image is separated into a binary image region and a multivalued image region based on a transformation result of a luminance signal by a discrete cosine transformation means. 5. The image encoding device according to claim 1, 2, 3, or 4, wherein the multilevel image encoding unit generates a binary image instructed from the outside instead of the area separated by the image separation unit. For each of the region and the multivalued image region, the binary image encoding unit performs two
An image encoding device characterized in that an image in a multivalued image region is encoded by a multivalued image coding unit. 6. The image encoding device according to claim 1, 2, 3, 4, or 5, wherein the multivalued image encoding section includes code word conversion means for converting the conversion result of the discrete cosine transformation means into code words. An image encoding device comprising: 7. The image encoding device according to claim 1, 2, 3, 4, 5, or 6, wherein the binary image encoding unit calculates a run length in a scanning line direction for an image in a binary image area. An image encoding device characterized by having a code word conversion means that performs code word conversion based on measurement. 8. The image encoding device according to claim 1, 2, 3, 4, 5, 6, or 7, characterized in that the image encoding device comprises another orthogonal transformation means in place of the discrete cosine transformation means. Encoding device. 9. A semiconductor integrated circuit for image processing, comprising the image encoding device according to claim 1, 2, 3, 4, 5, 6, 7, or 8. 10. Decoding the encoded binary image area image 2
An image decoding device comprising a value image decoding unit and a multi-value image decoding unit device that decodes an image in a multi-value image area encoded by discrete cosine transform, and claims 1, 2, 3, 4. ,
The image encoding device according to 5, 6, 7 or 8, or
An image encoding/decoding device comprising the semiconductor integrated circuit according to claim 9. 11. The image encoding/decoding device according to claim 10, wherein the image decoding device includes a conversion means for converting a multivalued image into a binary image. conversion device. 12. Claims 1, 2, 3, 4, 5 for encoding the input image.
, 6, 7, or 8, or the semiconductor integrated circuit according to claim 9, and communication control for controlling transmission of encoded image data encoded by the image encoding device or semiconductor integrated circuit provided therein. An image communication device characterized by having a section. 13. Claims 1, 2, 3, 4, 5 for encoding the input image.
, 6, 7, or 8, or the semiconductor integrated circuit according to claim 9, and a storage means for storing encoded image data encoded by the image encoding device or the semiconductor integrated circuit. An image storage device comprising: