JP3281183B2 - Image encoding / decoding method and image encoding / decoding device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image encoding / decoding method and an image encoding / decoding apparatus for encoding multi-valued image data in blocks of a predetermined shape.

[0002]

2. Description of the Related Art In recent years, there has been a demand for transmitting or storing multi-valued images obtained by reading with a scanner or the like as they are. In general, the amount of information of a multi-valued image is enormous, and it is impractical to transmit or accumulate it as it is. Therefore, various multi-valued image encoding methods capable of efficiently compressing a multi-valued image have been proposed.

For example, there is a so-called JPEG system as a system for directly encoding a multi-valued image. In the JPEG system, a multi-tone black-and-white image and a color image can be efficiently encoded and compressed.

On the other hand, with respect to binary images, a more efficient binary image coding system is required. That is, since the reading resolution of a scanner or the like has been greatly improved, the amount of image information per page has become enormous. Therefore, a binary image coding system which has been conventionally applied, for example, an MH coding system or an MR coding system In this case, a situation arises in which the amount of code data obtained after encoding does not become sufficiently small.

For this reason, a so-called JBIG binary coding system has been proposed and standardized.

[0006]

However, for example, some documents to be exchanged with a facsimile machine include a text document and a photo document mixed in one document. When the image data is encoded and compressed by the multi-level image encoding method, a character portion of an image obtained by decoding becomes unclear, and
If the image data is encoded and compressed only by the binary image encoding method, the gradation of the photograph portion of the decoded image is lost, and the image quality is degraded.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and it is possible to efficiently encode and compress an original image in which a character image and a photographic image are mixed, and to improve the image quality of an image obtained after decoding. It is an object of the present invention to provide an image encoding / decoding method and an image encoding / decoding device capable of improving the above.

[0008]

According to the present invention, there is provided an image encoding / decoding method for encoding multi-valued image data in block units of a predetermined shape. With respect to, while applying a prediction process of calculating the predicted pixel value of the target pixel based on the pixel values of a plurality of surrounding pixels adjacent to the target pixel,
Calculating a prediction error between the pixel value of the target pixel and the prediction pixel value, applying a predetermined statistical process to the prediction error, and applying a binary encoding process to the block based on the statistical processing result; , To apply multi-level encoding processing, or to determine whether to apply the encoding processing,
When it is determined that the binary coding process is applied, the binary code data obtained by applying a predetermined binary coding process to each pixel of the block and the code data are binary code data. Is output as code data of the block, and when it is determined that the multi-level encoding process is to be applied, a predetermined multi-level encoding process is applied to each pixel of the block. When the multi-level code data obtained thereby and the second identification code for identifying that the code data is multi-level code data are output as the code data of the block, and it is determined that the coding process is not applied, , A third identification code indicating that the encoding process has not been applied to the block is output as the code data of the block, while the first code is added to the code data at the time of decoding. When that contain identification code,
A predetermined binary decoding process is applied to the block to form original image data, and when the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block. To form the original image data, and when the code data includes the third identification code, execute the above-described prediction processing and form the image data in the block based on the prediction result. It was done.

Also, in an image coding / decoding method for coding multi-valued image data in block units of a predetermined shape, at the time of coding, each block is related to a target pixel to be coded and adjacent to the target pixel. A prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels is applied, and a prediction error between the pixel value of the target pixel and the predicted pixel value is calculated. Calculate the average value and the maximum value, classify the calculated average value and the maximum value based on a predetermined threshold, and apply a binary coding process to the block based on the classification result, It is determined whether to apply the coding process or not, and if it is determined that the binary coding process is to be applied, a predetermined binary coding process is applied to each pixel of the block. And outputs the binary code data obtained thereby and the first identification code for identifying that the code data is binary code data as code data of the block, and applying multi-level coding processing. If it is determined, a second multi-level code data is obtained by applying a predetermined multi-level coding process to each pixel of the block and identifying that the multi-level code data is multi-level code data. The code is output as the code data of the block, and when it is determined that the coding process is not applied, a third identification code indicating that the coding process is not applied to the block is used as the code data of the block. On the other hand, at the time of decoding, if the first identification code is included in the code data at the time of decoding, a predetermined binary decoding process is applied to the block to obtain the original image. When data is formed, and the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block to form original image data, and the third data is added to the code data. When the identification code is included, the above-described prediction processing is executed, and image data in the block is formed based on the prediction result.

Also, in an image encoding / decoding method for encoding multi-valued image data in block units of a predetermined shape, at the time of encoding, each block is related to a target pixel to be coded and is adjacent to the target pixel. A prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels is applied, and a prediction error between the pixel value of the target pixel and the predicted pixel value is calculated. Calculate the average value and the maximum value, classify the calculated average value and the maximum value based on a predetermined threshold, and apply a binary coding process to the block based on the classification result, or perform multi-level coding. It is determined whether to apply the encoding process or not to apply the encoding process. If it is determined that the binary encoding process is to be applied, a predetermined binary encoding process is applied to each pixel of the block. And outputs the binary code data obtained thereby and the first identification code for identifying that the code data is binary code data as code data of the block, and applying multi-level coding processing. If it is determined, a second multi-level code data is obtained by applying a predetermined multi-level coding process to each pixel of the block and identifying that the multi-level code data is multi-level code data. The code is output as the code data of the block, and when it is determined that the coding process is not applied, a third identification code indicating that the coding process is not applied to the block is used as the code data of the block. On the other hand, at the time of decoding, if the first identification code is included in the code data at the time of decoding, a predetermined binary decoding process is applied to the block to obtain the original image. When the second identification code is included in the code data, the original image data is formed by applying a predetermined multi-level decoding process to the block, and the third image is added to the code data. When the identification code is included, the prediction process is executed, image data in the block is formed based on the prediction result, and the predetermined threshold value can be changed from outside.

In an image coding / decoding method for coding multi-valued image data in block units having a predetermined shape, at the time of coding, each block is related to a target pixel to be coded and adjacent to the target pixel. A prediction process of calculating the predicted pixel value of the target pixel based on the pixel values of a plurality of surrounding pixels is applied, and a prediction error between the pixel value of the target pixel and the predicted pixel value is calculated. Apply predetermined statistical processing, and determine whether to apply binary coding processing, multi-level coding processing, or no coding processing for the block based on the statistical processing result. If it is determined that the binary encoding process is to be applied, a binary code obtained by applying a predetermined binary encoding process to the prediction error of each pixel of the block is obtained. The data and a first identification code for identifying that the code data is binary code data are output as the code data of the block, and when it is determined that the multi-level coding process is applied, each pixel of the block is output. A predetermined multi-level encoding process is applied to the prediction error of the multi-level code data obtained as a result and a second identification code for identifying that the code data is multi-level code data is used as code data of the block. Output, and when it is determined that the encoding process is not applied, a third identification code indicating that the encoding process is not applied to the block is output as the code data of the block. When the code data contains the first identification code, a predetermined binary decoding process is applied to the block to form a prediction error, and A prediction value is calculated by applying the above-described prediction processing to the block, and the calculated prediction error and the prediction value are combined to form the original image data. When the code data includes the second identification code, Applying a predetermined multi-level decoding process to the block to form a prediction error, applying a prediction process to the block to calculate a prediction value, and combining the calculated prediction error with the prediction value. When the original image data is formed, and the code data includes the third identification code, the prediction processing is executed, and the image data in the block is formed based on the prediction result. It is.

Also, in an image coding / decoding method for coding multi-valued image data in block units of a predetermined shape, at the time of coding, each block is related to a target pixel to be coded and adjacent to the target pixel. A prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels is applied, and a prediction error between the pixel value of the target pixel and the predicted pixel value is calculated. Calculate the average value and the maximum value, classify the calculated average value and the maximum value based on a predetermined threshold, and apply a binary coding process to the block based on the classification result, It is determined whether to apply the encoding process or not to apply the encoding process, and when it is determined to apply the binary encoding process, a predetermined value is set for the prediction error of each pixel of the block. A binary code data obtained by applying the binary coding process and a first identification code for identifying that the code data are binary code data are output as code data of the block, and a multi-level code is output. If it is determined that the coding process is applied, the multi-level code data obtained by applying a predetermined multi-level coding process to the prediction error of each pixel of the block and the code data are multi-level code data. A second identification code for identifying the presence of the block is output as code data of the block, and when it is determined that the coding process is not applied, a third code indicating that the coding process is not applied to the block is output. While outputting the identification code as code data of the block, at the time of decoding,
When the first identification code is included in the code data, a predetermined binary decoding process is applied to the block to form a prediction error, and the prediction value is applied to the block by applying the prediction process to the block. Calculate, combine the calculated prediction error and the predicted value to form the original image data, and when the code data includes the second identification code, perform a predetermined multi-level decoding process on the block. Applying the prediction error to form a prediction value, calculating the prediction value by applying the above-described prediction processing to the block, forming the original image data by combining the calculated prediction error and the prediction value, and When the third identification code is included, the above-described prediction processing is executed, and image data in the block is formed based on the prediction result.

In an image encoding / decoding method for encoding multi-valued image data in block units having a predetermined shape, at the time of encoding, for each block, a target pixel to be encoded is adjacent to the target pixel. A prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels is applied, and a prediction error between the pixel value of the target pixel and the predicted pixel value is calculated. Calculate the average value and the maximum value, classify the calculated average value and the maximum value based on a predetermined threshold, and apply a binary coding process to the block based on the classification result, It is determined whether to apply the encoding process or not to apply the encoding process, and when it is determined to apply the binary encoding process, a predetermined value is set for the prediction error of each pixel of the block. A binary code data obtained by applying the binary coding process and a first identification code for identifying that the code data are binary code data are output as code data of the block, and a multi-level code is output. If it is determined that the coding process is applied, the multi-level code data obtained by applying a predetermined multi-level coding process to the prediction error of each pixel of the block and the code data are multi-level code data. A second identification code for identifying the presence of the block is output as code data of the block, and when it is determined that the coding process is not applied, a third code indicating that the coding process is not applied to the block is output. While outputting the identification code as code data of the block, at the time of decoding,
When the first identification code is included in the code data, a predetermined binary decoding process is applied to the block to form a prediction error, and the prediction value is applied to the block by applying the prediction process to the block. Calculate, combine the calculated prediction error and the predicted value to form the original image data, and when the code data includes the second identification code, perform a predetermined multi-level decoding process on the block. Applying the prediction error to form a prediction value, calculating the prediction value by applying the above-described prediction processing to the block, forming the original image data by combining the calculated prediction error and the prediction value, and When the third identification code is included, the above-described prediction processing is executed, and image data in the block is formed based on the prediction result.

Also, in an image coding / decoding apparatus for coding multi-valued image data in block units of a predetermined shape, at the time of coding, each block is related to a target pixel to be coded and adjacent to the target pixel. Prediction processing means for calculating a predicted pixel value of the pixel of interest based on pixel values of a plurality of surrounding pixels; calculating a prediction error between the pixel value of the pixel of interest and the predicted pixel value; A determination unit that applies statistical processing and determines whether to apply binary coding processing, multi-level coding processing, or no coding processing to the block based on the result of the statistical processing. And when the determination means determines that the binary coding process is to be applied, the binary code obtained by applying the predetermined binary coding process to each pixel of the block. The data and a first identification code for identifying that the code data is binary code data are output as the code data of the block, and when it is determined that the multi-level coding process is applied, each pixel of the block is output. Applying a predetermined multi-level encoding process for the multi-level code data and a second identification code for identifying that the code data is multi-level code data are output as code data of the block, If it is determined that the encoding process is not applied, an image encoding unit that outputs a third identification code indicating that the encoding process is not applied to the block as code data of the block; ,
When the code data contains the first identification code, the block is subjected to a predetermined binary decoding process to form original image data, and the code data contains the second identification code. The original image data is formed by applying a predetermined multi-level decoding process to the block, and when the code data contains the third identification code, the prediction process is executed. It is provided with image decoding means for forming image data in the block based on the prediction result.

In an image coding / decoding apparatus for coding multi-valued image data in block units of a predetermined shape, at the time of coding, each block is related to a target pixel to be coded and is adjacent to the target pixel. Prediction processing means for applying a prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels, and calculating a prediction error between the pixel value of the target pixel and the predicted pixel value; Calculate the average value and the maximum value of the prediction error, classify the calculated average value and the maximum value based on a predetermined threshold, and apply the binary coding process to the block based on the classification result. Means for determining whether to apply multi-level coding processing or not to apply coding processing, and, when determining to apply binary coding processing, Applying a predetermined binary encoding process for the binary code data obtained thereby and a first identification code identifying that the code data is binary code data are output as code data of the block, If it is determined that the multi-level coding process is applied, the multi-level code data obtained by applying the predetermined multi-level coding process to each pixel of the block and the code data are multi-level code data. Is output as the code data of the block, and when it is determined that the encoding process is not applied, the third identification code indicating that the encoding process is not applied to the block is output. An image encoding means for outputting a code as code data of the block; and, when decoding, when the first identification code is included in the code data, decoding of the block. To form the original image data by applying a predetermined binary decoding process, and when the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block. Image decoding means for forming the original image data, executing the prediction processing when the code data includes the third identification code, and forming the image data in the block based on the prediction result It is provided with.

Also, in an image coding / decoding apparatus for coding multi-valued image data in block units of a predetermined shape, at the time of coding, for each block, a target pixel to be coded is adjacent to the target pixel. Prediction processing means for applying a prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels, and calculating a prediction error between the pixel value of the target pixel and the predicted pixel value; Calculate the average value and the maximum value of the prediction error, classify the calculated average value and the maximum value based on a predetermined threshold, and determine whether to apply the binary coding process to the block based on the classification result. Means for determining whether to apply multi-level coding processing or not to apply coding processing, and, when determining to apply binary coding processing, Applying a predetermined binary encoding process for the binary code data obtained thereby and a first identification code identifying that the code data is binary code data are output as code data of the block, If it is determined that the multi-level coding process is applied, the multi-level code data obtained by applying the predetermined multi-level coding process to each pixel of the block and the code data are multi-level code data. Is output as the code data of the block, and when it is determined that the encoding process is not applied, the third identification code indicating that the encoding process is not applied to the block is output. An image encoding means for outputting a code as code data of the block; and, when decoding, when the first identification code is included in the code data, decoding of the block. To form original image data by applying a predetermined binary decoding process, and when the code data includes the second identification code, applying a predetermined multi-level decoding process to the block. Image decoding means for forming the original image data, and when the code data contains the third identification code, executing the prediction processing and forming the image data in the block based on the prediction result And the predetermined threshold value applied by the determination means can be changed from outside.

Also, in an image coding / decoding apparatus for coding multi-valued image data in block units of a predetermined shape, at the time of coding, each block is related to a target pixel to be coded and adjacent to the target pixel. Prediction processing means for applying a prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels, and calculating a prediction error between the pixel value of the target pixel and the predicted pixel value; Applying a predetermined statistical process to the prediction error, applying a binary encoding process to the block based on the result of the statistical process, applying a multi-level encoding process,
Alternatively, a determination unit that determines whether to apply the encoding process, and when it is determined that the binary encoding process is to be applied,
Binary code data obtained by applying a predetermined binary coding process to the prediction error of each pixel of the block and a first identification code for identifying that the code data is binary code data Output as the code data of the block, and when it is determined that the multi-level coding process is to be applied, the multi-level code obtained by applying the predetermined multi-level coding process to the prediction error of each pixel of the block is obtained. When the data and a second identification code for identifying that the code data is multi-level code data are output as the code data of the block, and it is determined that the coding process is not applied,
An image encoding unit that outputs a third identification code indicating that the encoding process has not been applied to the block as code data of the block, and the first identification code is included in the code data during decoding. When a predetermined binary decoding process is applied to the block to form a prediction error, a prediction value is calculated by applying the prediction process to the block, and the calculated prediction error and prediction value are calculated. When the original image data is formed by synthesis, and when the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block to form a prediction error, and , A prediction value is calculated by applying the above-described prediction processing, and the calculated prediction error and the prediction value are combined to form original image data. When that contain different symbols, it performs the prediction process, in which includes the picture decoding means for forming an image data of the block based on the prediction result.

In an image coding / decoding apparatus for coding multi-valued image data in blocks of a predetermined shape, at the time of coding, for each block, a target pixel to be coded is adjacent to the target pixel. Prediction processing means for applying a prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels, and calculating a prediction error between the pixel value of the target pixel and the predicted pixel value; Calculate the average value and the maximum value of the prediction error, classify the calculated average value and the maximum value based on a predetermined threshold, and determine whether to apply the binary coding process to the block based on the classification result. Means for determining whether to apply multi-level coding processing or not to apply coding processing, and, when determining to apply binary coding processing, The binary code data obtained by applying a predetermined binary coding process to the above prediction error and a first identification code for identifying that the code data is binary code data are represented by code data of the block. And if it is determined that the multi-level encoding process is to be applied,
A multi-level code data obtained by applying a predetermined multi-level coding process to the prediction error of each pixel of the block and a second identification code for identifying that the code data is multi-level code data are obtained. Output as the coded data of the block, and when it is determined that the coding process is not applied,
An image encoding unit that outputs a third identification code indicating that the encoding process has not been applied to the block as code data of the block, and the first identification code is included in the code data during decoding. When a predetermined binary decoding process is applied to the block to form a prediction error, a prediction value is calculated by applying the prediction process to the block, and the calculated prediction error and prediction value are calculated. When the original image data is formed by synthesis, and when the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block to form a prediction error, and , A prediction value is calculated by applying the above-described prediction processing, and the calculated prediction error and the prediction value are combined to form original image data. When that contain different symbols, it performs the prediction process, in which includes the picture decoding means for forming an image data of the block based on the prediction result.

Further, in an image coding / decoding apparatus for coding multi-valued image data in block units of a predetermined shape, at the time of coding, each block is related to a target pixel to be coded and adjacent to the target pixel. Prediction processing means for applying a prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels, and calculating a prediction error between the pixel value of the target pixel and the predicted pixel value; Calculate the average value and the maximum value of the prediction error, classify the calculated average value and the maximum value based on a predetermined threshold, and determine whether to apply the binary coding process to the block based on the classification result. Means for determining whether to apply multi-level coding processing or not to apply coding processing, and, when determining to apply binary coding processing, The binary code data obtained by applying a predetermined binary coding process to the above prediction error and a first identification code for identifying that the code data is binary code data are represented by code data of the block. And if it is determined that the multi-level encoding process is to be applied,
A multi-level code data obtained by applying a predetermined multi-level coding process to the prediction error of each pixel of the block and a second identification code for identifying that the code data is multi-level code data are obtained. Output as the coded data of the block, and when it is determined that the coding process is not applied,
An image encoding unit that outputs a third identification code indicating that the encoding process has not been applied to the block as code data of the block, and the first identification code is included in the code data during decoding. When a predetermined binary decoding process is applied to the block to form a prediction error, a prediction value is calculated by applying the prediction process to the block, and the calculated prediction error and prediction value are calculated. When the original image data is formed by synthesis, and when the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block to form a prediction error, and , A prediction value is calculated by applying the above-described prediction processing, and the calculated prediction error and the prediction value are combined to form original image data. When that contain different symbols, performs the prediction process, includes the picture decoding means for forming an image data of the block based on the prediction result,
The predetermined threshold value applied by the determination means can be changed from outside.

In the prediction processing, when there is no adjacent pixel corresponding to the pixel of interest, a virtual pixel having a predetermined pixel value may be adopted as the adjacent pixel. Further, when the original image data is formed by applying the binary decoding process, the binary image data obtained by the binary decoding process may be converted into predetermined multi-valued image data.

[0021]

Therefore, since multi-level coding or binary coding is applied in units of blocks, character images and photographic images are appropriately coded, and the image quality of the image obtained after decoding is improved. . Further, when the pixels in a block can be generated by the prediction processing, the block is generated by the prediction processing at the time of decoding without generating the code data, so that the coding efficiency is greatly improved. . In addition, since conditions for judging an encoding process to be applied to each block can be appropriately set, a user can appropriately select a desired image quality or a processing speed. The performance is improved.

[0022]

Embodiments of the present invention will be described below in detail with reference to the accompanying drawings.

FIG. 1 shows a facsimile apparatus according to one embodiment of the present invention.

In FIG. 1, a system control unit 1 performs a control process of each unit of the facsimile apparatus and a facsimile transmission control procedure process. A system memory 2 stores a control processing program executed by the system control unit 1. , And various data required when executing the processing program, and constitutes a work area of the system control unit 1. The parameter memory 3 stores various information unique to the facsimile apparatus. It is for doing.

The multi-value scanner 4 reads a document image at a predetermined resolution in a multi-valued manner. The multi-value plotter 5 records and outputs an image at a predetermined resolution in a multi-valued manner. Is for operating the facsimile apparatus, and is composed of various operation keys and various displays.

The multi-level encoding / decoding section 7 applies a JPEG multi-level image encoding / decoding process for encoding a multi-level image in block units, and encodes and compresses the multi-level image signal. , For decoding the multi-valued image information that has been encoded and compressed into the original image signal.
Encodes a binary image signal by applying a binary image encoding / decoding process of the JBIG method that encodes the binary image signal in block units while binarizing the multivalued image signal. In addition, the binary image information that has been encoded and compressed is decoded into the original binary image signal, and the obtained binary image signal is converted into a predetermined multi-level image signal. .

The image storage device 9 is for storing a large number of image information in a coded and compressed state.
The facsimile modem 10 implements a group 3 facsimile modem function, and has a low-speed modem function (V.21 modem) for exchanging transmission procedure signals, and a high-speed modem mainly for exchanging image information. Functions (V.17 modem, V.29 modem, V.2
7 ter modem, etc.).

The network controller 11 is for connecting the facsimile apparatus to a public telephone line network and has an automatic transmission / reception function.

The system control unit 1, the system memory 2, the parameter memory 3, the multi-value scanner 4, the multi-value plotter 5, the operation display unit 6, the multi-value coding / decoding unit 7, the binary coding / decoding unit 8, an image storage device 9, a group 3 facsimile modem 10, and a network control device 11 are connected to an internal bus 12, and data exchange between these elements is mainly performed through the internal bus 12. Is being done.

Data exchange between the network controller 11 and the group 3 facsimile modem 10 is directly performed.

In this embodiment, multi-level image data obtained by reading with the multi-level scanner 4 is encoded and compressed by the following method.

First, as shown in FIG. 2A, the multivalued image data is divided into 8 × 8 size blocks BLK. Then, for all the pixels of this block BLK,
By applying a template TP as shown in FIG. 3B, a pixel prediction value of the target pixel X to be processed is calculated,
An error between the calculated pixel prediction value and the actual pixel value of the target pixel X (hereinafter, referred to as a prediction error) is calculated to form a prediction error for one block.

Here, the pixel prediction value is, for example, a simple value of the pixel value of the adjacent pixel A at the same pixel position on the line immediately before the target pixel X and the pixel value of the adjacent pixel B immediately before the same line on the target line X. The average ((A + B) / 2) is calculated and obtained. Also, for example, the coordinates in the block are (1, 1),
(1,2) (1,3), (1,4), (1,5),
(1,6) (1,7), (1,8) (2,1), (3,
1), (4,1), (5,1), (6,1), (7,
For the target pixel X of 1) and (8, 1) where the adjacent pixels A and B do not exist, a virtual value is applied as the value of the adjacent pixels A and B, respectively. As this virtual value, for example,
An intermediate value of the values that can be taken by the multi-valued image data can be used. As described above, the pixel values of the corresponding pixels in the adjacent block can be used as the adjacent pixels A and B of the target pixel X in which the adjacent pixels A and B do not exist in the same block.

Next, the average value DA and the maximum value DM of the prediction error values for one block are calculated, the statistical tendency of the prediction error of the block is classified, and multi-level image coding is applied to the block. It is determined whether to apply binary image coding or not to generate code data.

Here, for a block for which no code data has been created, multi-valued image data for one block is created by applying the above-described prediction operation at the time of decoding.
In this case, naturally, the multi-valued image data obtained by decoding includes an error. However, if the error is within an allowable range, the image quality of the block is small and the entire image of one page is reduced. The image quality degradation as described above can be reduced to a level that does not cause a problem. Further, the degree of the image quality deterioration can be appropriately set according to the setting condition of the judgment condition when classifying the statistical tendency of the prediction error.

After the encoding process, block code data as shown in FIG. 3A is formed for each block BLK. This block code data is obtained by adding a 2-bit flag indicating the type of code data of the block to the head of the code data of one block.

Then, as for the block which has not been subjected to the encoding process by the above-described encoding process, only the flag FN of the data "00" is arranged as the block code data as shown in FIG. .

For a block to which the binary image encoding process has been applied, as shown in FIG. 9C, a flag FA of data "01" and binary code data of the block are arranged in this order. Block code data is formed.

For a block to which the multi-level image encoding process has been applied, as shown in FIG. 11D, a flag FB of data "11" and multi-level code data of the block are arranged in this order. Block code data is formed.

FIG. 4 shows an example of processing at the time of encoding one page of a document image according to an embodiment of the present invention.

First, one page of the document image is read by the multi-value scanner 4, and the multi-value image data obtained thereby is temporarily stored in, for example, a read page buffer formed in the system memory 2. Is done.

In this state, one block of multivalued image data is extracted from the read page buffer (process 101).
A prediction error value for one block is calculated by applying the above-described prediction calculation (steps 102 and 103), and an average DA and a maximum value DM of the prediction error values are calculated (step 104).

Then, it is determined whether or not the average value DA is equal to or less than a predetermined threshold value R1 (decision 105).
When the result of Step 5 is YES, it is checked whether the maximum value DM is equal to or less than a predetermined threshold value R2 (decision 10).
6).

If the result of decision 106 is YES, the prediction error for that block is small overall and
Since the block does not include an image portion such as an edge portion of the image, in this case, it is determined that no code data is formed for the block. Then, the flag FN is stored in the code buffer as the code data of the block (process 107). Here, the code buffer is a buffer area formed in a predetermined area of the system memory 2.

When the result of decision 105 is NO or when the result of decision 106 is NO, code data is formed for that block. Therefore, it is checked whether or not the maximum value DM is larger than a predetermined threshold value R3 (> R2) (decision 108).

When the result of the judgment 108 is YES, it is a case that the image of the edge portion where the pixel value change of the pixel is very large is included in the block, and in this case, the image appears in the block. It is determined that the binary image coding is applied so that the image of the existing edge portion can be saved.

That is, in this case, the multivalued image data for one block of the block is encoded and compressed by the binary encoding / decoding section 8 (process 109), and the flag FA and the binary Encoding / decoding unit 8
Are sequentially stored in the code buffer (steps 110 and 111).

When the result of decision 108 is NO,
The block does not include the image of the edge portion. In such a case, it is determined that the multi-valued image coding is applied in order to improve the reproducibility of the image.

Therefore, in this case, one of the blocks
The multi-level image data for the block is encoded and compressed by the multi-level encoding / decoding unit 7 (process 112), and the flag FB and the multi-level code generated by the multi-level encoding / decoding unit 7 are used as the code data of the block. The data is sequentially stored in the code buffer (steps 113 and 114).

When the creation of the code data for one block and the storage in the code buffer are completed in this way, it is checked whether the processing for the multi-valued image data for one page has been completed at that time (determination). 115) When the result of the determination 115 is NO, the process returns to the process 101, and the encoding process of the remaining blocks is executed.

When the encoding process for one page has been completed and the result of determination 115 is YES, the encoded data for one page stored in the code buffer is output to the output destination designated at that time. (Process 11
6), this process ends. When the encoding process for one page is completed, the multi-valued image data of the processing target page stored in the page buffer is usually deleted.

Here, the values of the thresholds R1, R2, and R3 can be set to appropriate values according to the image quality specified for the original image to be read at that time. This designation is made by the user operating the operation display unit 6.

For example, the threshold value R1 is for checking the degree of the overall prediction error of the pixels for one block. If this value is set to a large value, the number of blocks in which code data is not formed increases. Although the data amount of the data is small, the deterioration of the image quality is large. Also, the threshold R
When the value of 1 is set to a small value, the number of blocks in which code data is not formed is reduced, so that the degree of image quality deterioration is reduced and the image quality is improved. The amount increases.

The threshold value R2 has a small prediction error,
This is for detecting a block including an edge image. Therefore, when this value is reduced, the ratio of detecting an edge image increases, and the number of blocks in which code data is not formed decreases. Conversely, when this value is increased, the ratio of detecting an edge image decreases, and the number of blocks in which code data is not formed increases.

The threshold value R3 is for determining whether or not a block to be encoded includes an edge image. Therefore, when this value is reduced, the ratio of detecting the edge image increases, and the image quality of the character portion image can be improved. Further, when this value is reduced, the rate of detecting the edge image is reduced, and the image quality of the image of the photograph can be further improved.

Therefore, for example, when the user designates "image quality priority; photograph", the number of blocks in which code data is not formed is reduced, and the number of blocks determined as a photographic image is increased. The threshold value R1 is reduced, the threshold value R2 is reduced, and the threshold value R3 is increased. Also,
When "image quality priority; text" is specified, the threshold value R1, the threshold value R2, and the threshold value R3 are reduced.

When "shortening of transmission time" is designated, the threshold R1 is increased, the threshold R2 is increased, and the threshold R2 is increased in order to increase the number of blocks in which code data is not formed.
Decrease 3 When the threshold value R3 is set to a small value, the number of blocks to be subjected to binary image encoding increases, and the amount of code data can be reduced.

Further, as the number of blocks in which code data is not formed increases, the encoding processing time for one page can be reduced. Therefore, by appropriately setting the values of the thresholds R1, R2, and R3, the encoding processing time can be reduced. Adjustment with image quality can be taken.

FIG. 5 shows an example of processing when decoding the code data for one page into the original multi-valued image data.

First, a 2-bit flag is extracted from the code data (processing 201), and the content of the flag is represented by a flag FN (= “0”).
0 ”), flag FA (=“ 01 ”), or flag FB (=“ 11 ”) (determination 202,
203).

When the extracted flag is the flag FN and the result of the judgment 202 is YES, the above-described prediction operation is applied to all the pixels of one block, and the predicted pixel value of one block formed thereby is obtained. Is stored in the block area of the page buffer corresponding to the block address of the current block (steps 204 and 20).
5). Here, the page buffer is for temporarily storing multi-valued image data for one page, and is formed in the system memory 2. The block address is address data for indicating an area where each block is arranged.

When the extracted flag is the flag FA and the result of determination 203 is YES, 1-bit code data is extracted, and the code data is transferred to the binary coding / decoding section 8 and A value image decoding process is executed (process 206), whereby the decoded data (multi-valued image data) output from the binary coding / decoding unit 8 is stored in the corresponding pixel address area of the block buffer (step 206). Processing 20
7). Here, the block buffer is for storing multi-valued image data for one block, and is formed in the system memory 2.

As described above, when the decoding process for one bit of the code data is completed, it is checked whether or not one block of multi-valued image data is ready in the block buffer (decision 208). Is N
When it becomes O, the process returns to the process 206 to execute the decoding process of the next code data.

When the multi-valued image data for one block has been prepared and the result of determination 208 is YES, the contents of the block buffer at that time are stored in the address space area of the page buffer corresponding to the block at that time. It is stored (process 209).

When the extracted flag is the flag FB and the result of determination 203 is NO, 1-bit code data is extracted, and the code data is transferred to the multi-level encoding / decoding unit 7 to Value image decoding processing (processing 2
10) As a result, the decoded data (multi-level image data) output from the multi-level encoding / decoding unit 7 is stored in the corresponding pixel address area of the block buffer (process 21).
1).

As described above, when the decoding process for one bit of the code data is completed, it is checked whether or not one block of multi-valued image data is ready in the block buffer (decision 212). Is N
When it becomes O, the process returns to the process 210 to execute the decoding process of the next code data.

When the multi-valued image data for one block has been prepared and the result of determination 212 is YES, the contents of the block buffer at that time are stored in the address space area of the page buffer corresponding to the block at that time. It is stored (process 213).

When the decoding process for one block is completed in this way, it is checked whether or not one page of multi-valued image data is stored in the page buffer at that time (decision 214). If the result is NO, the process returns to step 201 to execute the decoding process for the next block.

When the decoding process for one page has been completed and the result of determination 214 is YES, the multi-valued image data for one page stored in the page buffer at that time is (Step 215), and the image decoding process for one page is completed.

As described above, in the present embodiment, the coding method applied to the block is determined for each block, and the code data is not formed when there is no problem of the image quality deterioration. , The amount of code data can be significantly reduced.

The image quality is set by using threshold values R1, R2, R
By appropriately adjusting the value of 3, the value can be changed to some extent, so that the balance between the image quality required by the user and the coding efficiency can be appropriately balanced. It is preferable that the relationship between the image quality and the threshold values R1, R2, and R3 is determined in advance by an experiment, and a value corresponding to the user's image quality options is set.

FIG. 6 shows an example of processing for encoding an image of one page according to another embodiment of the present invention. Also in this case, similarly to the above-described embodiment, one page of the document image is read in advance by the multi-valued scanner 4, and
The multi-valued image data obtained thereby is temporarily stored in, for example, a read page buffer formed in the system memory 2.

In this state, one block of multi-valued image data is extracted from the read page buffer (process 301).
A prediction error value for one block is calculated by applying the above-described prediction calculation (steps 302 and 303), and an average DA and a maximum value DM of the prediction error values are calculated (step 304).

Then, it is checked whether or not the average value DA is equal to or less than a predetermined threshold value R1 (decision 305).
When the result of 5 is YES, the prediction error of the block is small as a whole, and in this case, it is determined that no code data is formed for the block.
Then, the flag FN is used as the code data of the block.
Is stored in the code buffer (process 306).

If the result of decision 305 is NO,
Code data is formed for the block. Therefore,
It is checked whether the maximum value DM is larger than a predetermined threshold value R2 (decision 307).

If the result of determination 307 is YES, it means that the image of the edge portion where the pixel value change of the pixel is very large is included in the block, and in this case, the image appears in the block. It is determined that the binary image coding is applied so that the image of the edge portion that exists can be stored.

That is, in this case, the multivalued image data for one block of the block is encoded and compressed by the binary encoding / decoding section 8 (process 308), and the flag FA and the binary Encoding / decoding unit 8
Are sequentially stored in the code buffer (steps 309 and 310).

If the result of decision 307 is NO,
The block does not include the image of the edge portion. In such a case, it is determined that the multi-valued image coding is applied in order to improve the reproducibility of the image.

Therefore, in this case, one of the blocks
The multi-level image data for the block is encoded and compressed by the multi-level encoding / decoding unit 7 (process 311), and the flag FB and the multi-level code formed by the multi-level encoding / decoding unit 7 are used as the code data of the block. The data is sequentially stored in the code buffer (processes 312 and 313).

When the creation of the code data for one block and the storage in the code buffer are completed in this way, it is checked whether the processing for the multi-valued image data for one page has been completed at that time (determination). 314) When the result of the determination 314 is NO, the process returns to the process 301 to execute the encoding process of the remaining blocks.

When the encoding process for one page is completed and the result of determination 314 is YES, the encoded data for one page stored in the code buffer is output to the output destination designated at that time. Output to (process 31
5), this process ends. When the encoding process for one page is completed, the multi-valued image data of the processing target page stored in the page buffer is usually deleted.

In this case, it is determined whether or not to form the code data of the block using only the threshold value R1.
In this case, the same effect as described above can be obtained. However,
In this case, since the prediction error of the entire block is small and the encoding process is not performed on the block including the edge image, the image quality of the character portion may be deteriorated accordingly.

In the above embodiment, the multi-valued image data is directly encoded. However, depending on the contents of the image, encoding the prediction error value improves the encoding efficiency. There is.

FIG. 7 shows a case where a prediction error value is encoded.
An example of encoding processing for a page is shown. Also in this case, similarly to the above-described embodiment, the multi-valued scanner 4 reads a document image for one page in advance, and the multi-valued image data obtained thereby is formed in the system memory 2, for example. Is temporarily stored in the read page buffer.

In this state, one block of multi-valued image data is extracted from the read page buffer (process 401).
By applying the above-described prediction calculation, a prediction error value for one block is calculated (processes 402 and 403), and an average DA and a maximum value DM of the prediction error values are calculated (process 404).

Then, it is determined whether or not the average value DA is equal to or smaller than a predetermined threshold value R1 (decision 405).
If the result of Step 5 is YES, it is checked whether or not the maximum value DM is equal to or less than a predetermined threshold value R2 (decision 40).
6).

If the result of decision 406 is YES, the prediction error for that block is small overall and
Since the block does not include an image portion such as an edge portion of the image, in this case, it is determined that no code data is formed for the block. Then, the flag FN is stored in the code buffer as the code data of the block (process 407). Here, the code buffer is a buffer area formed in a predetermined area of the system memory 2.

When the result of decision 405 is NO, or when the result of decision 406 is NO, code data is formed for that block. Then, it is checked whether or not the maximum value DM is larger than a predetermined threshold value R3 (> R2) (decision 408).

If the result of decision 408 is YES, it means that the image of the edge portion where the pixel value change of the pixel is very large is included in the block, and in this case, the image appears in the block. It is determined that the binary image coding is applied so that the image of the existing edge portion can be saved.

That is, in this case, the prediction error value for one block of the block is coded and compressed by the binary coding / decoding section 8 (processing 409), and the flag FA and the binary code The binary code data formed by the decoding / decoding unit 8 is sequentially stored in a code buffer (steps 410 and 411).

When the result of decision 408 is NO,
The block does not include the image of the edge portion. In such a case, it is determined that the multi-valued image coding is applied in order to improve the reproducibility of the image.

Therefore, in this case, one of the blocks
The prediction error value for the block is encoded and compressed by the multi-level encoding / decoding section 7 (processing 412), and the flag FB and the multi-level code data formed by the multi-level encoding / decoding section 7 are used as the code data of the block. Are sequentially stored in the code buffer (processes 413 and 414).

When the creation of the code data for one block and the storage in the code buffer are completed in this way, it is checked whether or not the processing for the multi-valued image data for one page has been completed at that time (determination). 415) When the result of the determination 415 is NO, the process returns to the process 401 to execute the encoding process of the remaining blocks.

When the encoding process for one page is completed and the result of determination 415 is YES, the encoded data for one page stored in the code buffer is output to the output destination designated at that time. (Process 41)
6), this process ends. When the encoding process for one page is completed, the multi-valued image data of the processing target page stored in the page buffer is usually deleted.

FIG. 8 shows an example of a process for decoding the code data for one page formed in this way.

First, a 2-bit flag is extracted from the code data (process 501), and the content of the flag is determined by the flag FN (= “0”).
0 "), flag FA (=" 01 "), or flag FB (=" 11 ") (determination 502,
503).

When the extracted flag is the flag FN and the result of the decision 502 is YES, the above-described prediction operation is applied to all the pixels of one block, and the predicted pixel value of one block formed thereby is obtained. Is stored in the block area of the page buffer corresponding to the block address of the block at that time (processes 504 and 50).
5). Here, the page buffer is for temporarily storing multi-valued image data for one page, and is formed in the system memory 2. The block address is address data for indicating an area where each block is arranged.

If the extracted flag is the flag FA and the result of decision 503 is YES, 1-bit code data is extracted, and the code data is transferred to the binary coding / decoding section 8 and A value image decoding process is executed (process 506), whereby the decoded data (prediction error value (multi-valued image data)) output from the binary coding / decoding unit 8
Is stored in the corresponding pixel address area of the block buffer (process 507). Here, the block buffer is
This is for storing one block of multivalued image data, and is formed in the system memory 2.

When the decoding process for one bit of the code data is completed in this way, it is checked whether or not the prediction error values for one block are completed in the block buffer at this time (decision 508). When the result is NO, the process returns to the process 506 to execute the decoding process of the code data of the next bit.

When one block of multi-valued image data has been prepared and the result of decision 508 is YES, the above-described prediction operation is applied to all pixels of one block stored in the block buffer, and A predicted pixel value for one block is calculated (step 509), and the formed predicted pixel value is combined with a prediction error stored in a block buffer to form multi-valued image data for one block. The contents of the block buffer are rewritten with the contents of the value image data (process 510). Then, the contents of the block buffer are stored in the address space area of the page buffer corresponding to the block at that time (process 511).

If the extracted flag is the flag FB and the result of determination 503 is NO, 1-bit code data is extracted and the Value image decoding processing (processing 5
12) Thus, the decoded data (prediction error value) output from the multi-level encoding / decoding unit 7 is stored in the corresponding pixel address area of the block buffer (process 513).

When the decoding process for one bit of the code data is completed in this way, at this time, it is checked whether or not the prediction error values for one block are complete in the block buffer (decision 514). When the result is NO, the process returns to the process 512 to execute the decoding process of the next code data.

When the multi-valued image data for one block has been prepared and the result of decision 514 is YES, the above-described prediction operation is applied to all the pixels for one block stored in the block buffer. A predicted pixel value for one block is calculated (step 515), and the formed predicted pixel value and the prediction error stored in the block buffer are combined to form multi-valued image data for one block. The contents of the block buffer are rewritten with the contents of the value image data (process 516). Then, the contents of the block buffer are stored in the address space area of the page buffer corresponding to the block at that time (process 517).

As described above, when the decoding process for one block is completed, it is checked whether or not one page of multi-valued image data is stored in the page buffer at that time (decision 518). If the result is NO, the process returns to step 501 to execute the decoding process for the next block.

In the above-described embodiment, a code indicating the end of the block is not added to the end of the code data of each block, but a code indicating the end of the block may be added. In this case, the end of one block of code data can be determined with reference to the block end code, so that the decoding end determination process for one block can be further simplified.

However, in an encoding system such as the JPEG system or JBIG system of the above-described embodiment, in which an image is originally encoded in image units divided into blocks, the end of the block is added to the end of the code data. In some cases, a code representing the code data is added or a code indicating the length of the code data of one block is added to the leading end of the code data. Can be determined.

In the above-described embodiment, the case where the JPEG system is applied as the multi-valued image coding system and the JBIG system is applied as the binary image coding system has been described. However, other coding systems are used. You can also. For example, as the binary image coding method, an MH coding method, an MR coding method, or an MMR coding method can be used.

In the above-described embodiment, the present invention is applied to a facsimile apparatus. However, the present invention can be applied to other apparatuses in the same manner.

[0109]

As described above, according to the present invention,
Since the multi-level encoding process or the binary encoding process is applied for each block, the character image and the photographic image are appropriately encoded, and the image quality obtained after decoding is improved.

When the pixels in a block can be generated by the prediction process, the block is generated by the prediction process at the time of decoding without generating the code data. It also has the effect of improving.

Further, since conditions for determining the encoding process to be applied to each block can be appropriately set, the user can appropriately select the desired image quality or processing speed, and The effect that the usability of the processing is improved is also obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a facsimile apparatus according to one embodiment of the present invention.

FIG. 2 is a schematic diagram for explaining an outline of an encoding process.

FIG. 3 is a schematic diagram showing an example of a signal format of code data.

FIG. 4 is a flowchart showing an example of encoding processing of an image for one page according to an embodiment of the present invention.

FIG. 5 is a flowchart illustrating an example of a decoding process for one page according to an embodiment of the present invention.

FIG. 6 is a flowchart showing an image encoding process for one page according to another embodiment of the present invention.

FIG. 7 is a flowchart showing encoding processing of an image for one page according to still another embodiment of the present invention.

FIG. 8 is a flowchart showing a decoding process for one page according to still another embodiment of the present invention.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 system control unit 2 system memory 7 multi-level encoding / decoding unit 8 binary encoding / decoding unit

Claims (16)

(57) [Claims] 1. An image coding / decoding method for coding multi-valued image data in units of a block having a predetermined shape, wherein, at the time of coding, each block is related to a target pixel to be coded and adjacent to the target pixel. A prediction process of calculating the predicted pixel value of the target pixel based on the pixel values of a plurality of surrounding pixels is applied, and a prediction error between the pixel value of the target pixel and the predicted pixel value is calculated. Apply predetermined statistical processing, and determine whether to apply binary coding processing, multi-level coding processing, or no coding processing for the block based on the statistical processing result. If it is determined that the binary coding process is to be applied, binary code data obtained by applying a predetermined binary coding process to each pixel of the block and its code A first identification code for identifying that the data is binary code data is output as code data of the block, and when it is determined that the multi-level encoding process is to be applied, a predetermined value is determined for each pixel of the block. Applying multi-level coding processing to output multi-level code data obtained by the multi-level coding processing and a second identification code for identifying the code data as multi-level code data as code data of the block; If it is determined that the block is not applied, a third identification code indicating that the coding process has not been applied to the block is output as the code data of the block, while, at the time of decoding, the first data is added to the code data at the time of decoding. When the block includes the identification code, the original image data is formed by applying a predetermined binary decoding process to the block, and the code data includes the second identification code. If it is, the predetermined image is applied to the block to form the original image data, and if the code data includes the third identification code, the prediction process is executed. An image encoding / decoding method characterized by forming image data in the block based on the prediction result. 2. An image coding / decoding method for coding multi-valued image data in units of a block having a predetermined shape, wherein at the time of coding, a target pixel to be coded is adjacent to the target pixel for encoding for each block. A prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels is applied, and a prediction error between the pixel value of the target pixel and the predicted pixel value is calculated. Calculate the average value and the maximum value, classify the calculated average value and the maximum value based on a predetermined threshold, and apply a binary coding process to the block based on the classification result, It is determined whether to apply the encoding process or not to apply the encoding process, and when it is determined to apply the binary encoding process, a predetermined binary encoding process is performed for each pixel of the block. The binary code data obtained thereby and the first identification code for identifying that the code data is binary code data are output as the code data of the block, and it is determined that the multi-level coding process is applied. In this case, the multi-level code data obtained by applying a predetermined multi-level coding process to each pixel of the block and a second identification code for identifying that the code data is multi-level code data Is output as the code data of the block, and when it is determined that the coding process is not applied, a third identification code indicating that the coding process is not applied to the block is output as the code data of the block. On the other hand, at the time of decoding, when the first identification code is included in the code data, a predetermined binary decoding process is applied to the block to obtain the original image data. When the second identification code is included in the code data, a predetermined multi-level decoding process is applied to the block to form original image data, and the third data is added to the code data. An image encoding / decoding method, characterized in that when the identification code is included, the prediction processing is performed, and image data in the block is formed based on the prediction result. 3. An image encoding / decoding method for encoding multi-valued image data in block units of a predetermined shape, wherein at encoding, for each block, a pixel of interest to be encoded is adjacent to the pixel of interest. A prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels is applied, and a prediction error between the pixel value of the target pixel and the predicted pixel value is calculated. Calculate the average value and the maximum value, classify the calculated average value and the maximum value based on a predetermined threshold, and apply a binary coding process to the block based on the classification result, or perform multi-level coding. It is determined whether to apply the encoding process or not to apply the encoding process, and when it is determined to apply the binary encoding process, a predetermined binary encoding process is performed for each pixel of the block. The binary code data obtained thereby and the first identification code for identifying that the code data is binary code data are output as the code data of the block, and it is determined that the multi-level coding process is applied. In this case, the multi-level code data obtained by applying a predetermined multi-level coding process to each pixel of the block and a second identification code for identifying that the code data is multi-level code data Is output as the code data of the block, and when it is determined that the coding process is not applied, a third identification code indicating that the coding process is not applied to the block is output as the code data of the block. On the other hand, at the time of decoding, when the first identification code is included in the code data, a predetermined binary decoding process is applied to the block to obtain the original image data. When the second identification code is included in the code data, a predetermined multi-level decoding process is applied to the block to form original image data, and the third data is added to the code data. When the identification code is included, the image processing device executes the prediction process, forms image data in the block based on the prediction result, and enables the predetermined threshold value to be changed from outside. Encoding / decoding method. 4. An image coding / decoding method for coding multi-valued image data in units of a block having a predetermined shape, wherein at the time of coding, for each block, a target pixel to be coded is adjacent to the target pixel. A prediction process of calculating the predicted pixel value of the target pixel based on the pixel values of a plurality of surrounding pixels is applied, and a prediction error between the pixel value of the target pixel and the predicted pixel value is calculated. Apply predetermined statistical processing, and determine whether to apply binary coding processing, multi-level coding processing, or no coding processing for the block based on the statistical processing result. If it is determined that the binary coding process is to be applied, a predetermined binary coding process is applied to the prediction error of each pixel of the block to obtain a binary code data. And outputs a first identification code for identifying that the code data is binary code data as the code data of the block, and when it is determined that the multi-level coding process is applied, each pixel of the block is output. A predetermined multi-level encoding process is applied to the prediction error of the multi-level code data obtained as a result and a second identification code for identifying that the code data is multi-level code data is used as code data of the block. Output, and when it is determined that the encoding process is not applied, a third identification code indicating that the encoding process is not applied to the block is output as the code data of the block. When the code data includes the first identification code, a predetermined binary decoding process is applied to the block to form a prediction error, and By applying the prediction process to calculate a predicted value for the click,
The calculated prediction error and prediction value are combined to form original image data. When the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block. A prediction error is formed, a prediction value is calculated by applying the above-described prediction processing to the block, the calculated prediction error and the prediction value are combined to form original image data, and the third image is added to the code data. An image encoding / decoding method, characterized in that when the identification code is included, the prediction processing is performed, and image data in the block is formed based on the prediction result. 5. An image coding / decoding method for coding multi-valued image data in block units having a predetermined shape, wherein at the time of coding, for each block, a target pixel to be coded is adjacent to the target pixel. A prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels is applied, and a prediction error between the pixel value of the target pixel and the predicted pixel value is calculated. Calculate the average value and the maximum value, classify the calculated average value and the maximum value based on a predetermined threshold, and apply a binary coding process to the block based on the classification result, or perform multi-level coding. It is determined whether to apply the encoding process or not to apply the encoding process. If it is determined that the binary encoding process is to be applied, the prediction error of each pixel of the block is determined by a predetermined value. Applying a coding process, outputting binary code data obtained by the coding process and a first identification code for identifying that the code data is binary code data, as code data of the block; If it is determined that is applied, multi-level code data obtained by applying a predetermined multi-level coding process to the prediction error of each pixel of the block and the code data are multi-level code data Is output as the code data of the block, and when it is determined that the encoding process is not applied, the third identification code indicating that the encoding process is not applied to the block Is output as the code data of the block. On the other hand, at the time of decoding, when the first identification code is included in the code data, a predetermined binary decoding is performed on the block. To form the prediction error by applying a process to calculate a predicted value by applying the prediction process for the block,
The calculated prediction error and prediction value are combined to form original image data. When the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block. A prediction error is formed, a prediction value is calculated by applying the above-described prediction processing to the block, the calculated prediction error and the prediction value are combined to form original image data, and the third image is added to the code data. An image encoding / decoding method, characterized in that when the identification code is included, the prediction processing is performed, and image data in the block is formed based on the prediction result. 6. An image coding / decoding method for coding multi-valued image data in block units having a predetermined shape, wherein each block includes a target pixel to be coded adjacent to the target pixel at the time of coding. A prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels is applied, and a prediction error between the pixel value of the target pixel and the predicted pixel value is calculated. Calculate the average value and the maximum value, classify the calculated average value and the maximum value based on a predetermined threshold, and apply a binary coding process to the block based on the classification result, or perform multi-level coding. It is determined whether to apply the encoding process or not to apply the encoding process. If it is determined that the binary encoding process is to be applied, the prediction error of each pixel of the block is determined by a predetermined value. Applying a coding process, outputting binary code data obtained by the coding process and a first identification code for identifying that the code data is binary code data, as code data of the block; If it is determined that is applied, multi-level code data obtained by applying a predetermined multi-level coding process to the prediction error of each pixel of the block and the code data are multi-level code data Is output as the code data of the block, and when it is determined that the encoding process is not applied, the third identification code indicating that the encoding process is not applied to the block Is output as the code data of the block. On the other hand, at the time of decoding, when the first identification code is included in the code data, a predetermined binary decoding is performed on the block. To form the prediction error by applying a process to calculate a predicted value by applying the prediction process for the block,
The calculated prediction error and prediction value are combined to form original image data. When the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block. A prediction error is formed, a prediction value is calculated by applying the above-described prediction processing to the block, and the calculated prediction error and the prediction value are combined to form original image data. When the identification code is included, the image processing device executes the prediction process, forms image data in the block based on the prediction result, and enables the predetermined threshold value to be changed from outside. Encoding / decoding method. 7. The prediction process according to claim 1, wherein when there is no adjacent pixel corresponding to the target pixel, a virtual pixel having a predetermined pixel value is adopted as an adjacent pixel. Or claim 3 or claim 4
7. The image encoding / decoding method according to claim 5 or claim 6. 8. When applying the binary decoding process to form original image data, converting the binary image data obtained by the binary decoding process into predetermined multi-valued image data. 4. The image encoding / decoding method according to claim 1, wherein the image encoding / decoding method is performed. 9. An image coding / decoding apparatus for coding multi-valued image data in block units having a predetermined shape, wherein each block is associated with a target pixel to be coded and adjacent to the target pixel at the time of coding. Prediction processing means for calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels; calculating a prediction error between the pixel value of the target pixel and the predicted pixel value; A determination unit that applies statistical processing and determines whether to apply binary coding processing, multi-level coding processing, or no coding processing to the block based on the result of the statistical processing. When the determination means determines that the binary coding process is to be applied, the predetermined binary coding process is applied to each pixel of the block, and the binary code data obtained thereby is obtained. Data and the first identification code for identifying that the code data is binary code data are output as the code data of the block, and when it is determined that the multi-level coding process is to be applied, each of the blocks of the block is output. A predetermined multi-level coding process is applied to the pixel, and multi-level code data obtained thereby and a second identification code for identifying that the code data is multi-level code data are output as code data of the block. If it is determined that the encoding process is not applied, an image encoding unit that outputs a third identification code indicating that the encoding process is not applied to the block as code data of the block, When the first identification code is included in the code data, the original image data is formed by applying a predetermined binary decoding process to the block, and the code data is generated. When the second identification code is included in the data, the original image data is formed by applying a predetermined multi-level decoding process to the block, and the code data includes the third identification code. An image encoding / decoding apparatus comprising: an image decoding unit that executes the above-described prediction processing and forms image data in the block based on the prediction result. 10. An image encoding / decoding apparatus which encodes multi-valued image data in block units having a predetermined shape, wherein at the time of encoding, for each block, a target pixel to be encoded is adjacent to the target pixel. Prediction processing means for applying a prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels, and calculating a prediction error between the pixel value of the target pixel and the predicted pixel value; Calculate the average value and the maximum value of the prediction error, classify the calculated average value and the maximum value based on a predetermined threshold, and determine whether to apply the binary coding process to the block based on the classification result. A determination means for determining whether to apply the multi-level encoding process or not to apply the encoding process, and, if it is determined to apply the binary encoding process, each pixel of the block. Applying a predetermined binary encoding process for the binary code data obtained thereby and a first identification code identifying that the code data is binary code data are output as code data of the block, If it is determined that the multi-level coding process is applied, the multi-level code data obtained by applying the predetermined multi-level coding process to each pixel of the block and the code data are multi-level code data. Is output as the code data of the block, and when it is determined that the encoding process is not applied, the third identification code indicating that the encoding process is not applied to the block is output. An image encoding means for outputting a code as code data of the block; and, when decoding, when the first identification code is included in the code data, the image coding means outputs the code. To form the original image data by applying a predetermined binary decoding process, and when the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block. Image decoding means for forming the original image data, executing the prediction processing when the code data includes the third identification code, and forming the image data in the block based on the prediction result An image encoding / decoding device comprising: 11. An image coding / decoding apparatus for coding multi-valued image data in block units of a predetermined shape, wherein each block is associated with a target pixel to be coded and adjacent to the target pixel at the time of coding. Prediction processing means for applying a prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels, and calculating a prediction error between the pixel value of the target pixel and the predicted pixel value; Calculate the average value and the maximum value of the prediction error, classify the calculated average value and the maximum value based on a predetermined threshold, and determine whether to apply the binary coding process to the block based on the classification result. A determination means for determining whether to apply the multi-level encoding process or not to apply the encoding process, and, if it is determined to apply the binary encoding process, each pixel of the block. Applying a predetermined binary encoding process for the binary code data obtained thereby and a first identification code identifying that the code data is binary code data are output as code data of the block, If it is determined that the multi-level coding process is applied, the multi-level code data obtained by applying the predetermined multi-level coding process to each pixel of the block and the code data are multi-level code data. Is output as the code data of the block, and when it is determined that the encoding process is not applied, the third identification code indicating that the encoding process is not applied to the block is output. An image encoding means for outputting a code as code data of the block; and, when decoding, when the first identification code is included in the code data, the image coding means outputs the code. To form original image data by applying a predetermined binary decoding process, and when the code data includes the second identification code, applying a predetermined multi-level decoding process to the block. Image decoding means for forming the original image data, and when the code data contains the third identification code, executing the prediction processing and forming the image data in the block based on the prediction result An image encoding / decoding apparatus, characterized in that a predetermined threshold value applied by the determination means can be changed from outside. 12. An image encoding / decoding apparatus which encodes multi-valued image data in a block unit having a predetermined shape, wherein at the time of encoding, for each block, a target pixel to be encoded is adjacent to the target pixel. Prediction processing means for applying a prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels, and calculating a prediction error between the pixel value of the target pixel and the predicted pixel value; A predetermined statistical process is applied to the prediction error, and based on the result of the statistical process, a binary coding process is applied to the block, a multi-level coding process is applied, or no coding process is applied. Determining means for determining whether or not to apply a binary coding process, and applying a predetermined binary coding process to the prediction error of each pixel of the block when determining that the binary coding process is to be applied; Then, the obtained binary code data and the first identification code for identifying the code data as binary code data are output as the code data of the block, and it is determined that the multi-level coding process is applied. In this case, the second multi-level code data is obtained by applying a predetermined multi-level coding process to the prediction error of each pixel of the block and identifying that the multi-level code data is multi-level code data. The identification code is output as the code data of the block, and if it is determined that the coding process is not applied, a third identification code indicating that the coding process is not applied to the block is code data of the block. Image decoding means for outputting as the above, and when decoding, when the first identification code is included in the code data, a predetermined binary decoding process is performed on the block. Apply to form a prediction error, calculate the predicted value by applying the above prediction process for the block,
The calculated prediction error and prediction value are combined to form original image data. When the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block. A prediction error is formed, a prediction value is calculated by applying the above-described prediction processing to the block, the calculated prediction error and the prediction value are combined to form original image data, and the third image is added to the code data. An image encoding / decoding apparatus comprising: an image decoding unit that executes the prediction processing when an identification code is included and forms image data in the block based on the prediction result. 13. An image encoding / decoding apparatus which encodes multi-valued image data in block units of a predetermined shape, wherein at the time of encoding, for each block, a target pixel to be encoded is adjacent to the target pixel. Prediction processing means for applying a prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels, and calculating a prediction error between the pixel value of the target pixel and the predicted pixel value; Calculate the average value and the maximum value of the prediction error, classify the calculated average value and the maximum value based on a predetermined threshold, and determine whether to apply the binary coding process to the block based on the classification result. A determination means for determining whether to apply the multi-level encoding process or not to apply the encoding process, and, if it is determined to apply the binary encoding process, each pixel of the block. The binary code data obtained by applying a predetermined binary coding process to the above prediction error and a first identification code for identifying that the code data is binary code data are represented by code data of the block. And if it is determined that the multi-level coding process is to be applied, multi-level code data obtained by applying a predetermined multi-level coding process to the prediction error of each pixel of the block and its code A second identification code for identifying that the data is multi-level code data is output as the code data of the block, and when it is determined that the coding process is not applied, the coding process is applied to the block. Image encoding means for outputting a third identification code indicating that there is no data as the code data of the block, and the first identification code is included in the code data upon decoding. Sometimes, a prediction error is formed by applying a predetermined binary decoding process to the block, and a prediction value is calculated by applying the prediction process to the block,
The calculated prediction error and prediction value are combined to form original image data. When the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block. A prediction error is formed, a prediction value is calculated by applying the above-described prediction processing to the block, the calculated prediction error and the prediction value are combined to form original image data, and the third image is added to the code data. An image encoding / decoding apparatus comprising: an image decoding unit that executes the prediction processing when an identification code is included and forms image data in the block based on the prediction result. 14. An image coding / decoding apparatus for coding multi-valued image data in units of a block having a predetermined shape, wherein at the time of coding, for each block, a target pixel to be coded is adjacent to the target pixel. Prediction processing means for applying a prediction process of calculating a predicted pixel value of the target pixel based on pixel values of a plurality of surrounding pixels, and calculating a prediction error between the pixel value of the target pixel and the predicted pixel value; Calculate the average value and the maximum value of the prediction error, classify the calculated average value and the maximum value based on a predetermined threshold, and determine whether to apply the binary coding process to the block based on the classification result. A determination means for determining whether to apply the multi-level encoding process or not to apply the encoding process, and, if it is determined to apply the binary encoding process, each pixel of the block. The binary code data obtained by applying a predetermined binary coding process to the above prediction error and a first identification code for identifying that the code data is binary code data are represented by code data of the block. And if it is determined that the multi-level coding process is to be applied, multi-level code data obtained by applying a predetermined multi-level coding process to the prediction error of each pixel of the block and its code A second identification code for identifying that the data is multi-level code data is output as the code data of the block, and when it is determined that the coding process is not applied, the coding process is applied to the block. Image encoding means for outputting a third identification code indicating that there is no data as the code data of the block, and the first identification code is included in the code data upon decoding. Sometimes, a prediction error is formed by applying a predetermined binary decoding process to the block, and a prediction value is calculated by applying the prediction process to the block,
The calculated prediction error and prediction value are combined to form original image data. When the code data includes the second identification code, a predetermined multi-level decoding process is applied to the block. A prediction error is formed, a prediction value is calculated by applying the above-described prediction processing to the block, the calculated prediction error and the prediction value are combined to form original image data, and the third image is added to the code data. When the identification code is included, the image processing apparatus includes image decoding means for executing the prediction processing and forming image data in the block based on the prediction result. An image encoding / decoding device characterized in that the image encoding / decoding device can be changed from the above. 15. The prediction process according to claim 9, wherein when there is no adjacent pixel corresponding to the target pixel, a virtual pixel having a predetermined pixel value is adopted as an adjacent pixel. An image encoding / decoding apparatus according to claim 11, claim 12, claim 13, or claim 14. 16. When forming the original image data by applying the binary decoding process, converting the binary image data obtained by the binary decoding process into predetermined multi-valued image data. Claim 9 or Claim 10 or Claim 1
2. The image encoding / decoding device according to 1.