JP3439933B2 - Code processing device for multi-valued image data - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a multi-valued image data code processing device, for example, an encoding method for reducing the amount of memory used when reproducing the multi-valued image data after the multi-valued image data is encoded or reproduced. Regarding what to decrypt.

[0002]

2. Description of the Related Art With the recent increase in speed and sophistication of data processing, monotone or color multi-gradation images are read or created on a display and transmitted, for example, between facsimile machines or personal computers. Processing and storage processing in a data storage device are performed. However, since the data amount of multi-valued image data of a multi-gradation image is an order of magnitude larger than that of the conventional black and white binary image data, the communication cost will be enormous if it is transmitted as it is, and a large-capacity storage means is required to store it. Will be required. Therefore, it is general that the multi-valued image data is compression-encoded and then subjected to transmission processing and storage processing.

A code processing device for multi-valued image data of this type is described in, for example, Japanese Patent Laid-Open No. 5-300382. This code processing device employs a so-called bit plane coding system, and is configured to expand multi-valued image data into bit planes before compression processing and compress and code each plane data. In the transmission by the bit plane encoding method, data is encoded and transmitted for each plane in order from the upper bit (MSB) side or the lower bit (LSB) side of the bit plane.

[0004]

However, in such a conventional multi-valued image data code processing apparatus,
When transmitting a high quality image, for example, an image of about 8 bits / pixel, 8 bit planes are required to form one image, and even if each plane is coded and compressed, 8 It takes communication time for the bit plane. As a method of reducing the number of bit planes, there is a method that uses a multi-valued dither method. This multi-valued dither method has the effect of reducing the amount of data while suppressing the deterioration of image quality. There is a problem that the above data becomes discrete and the compression rate decreases. As a result, the communication time may not be shortened as expected.

Here, in order to improve the compression rate compared with the multi-valued dither method, when the multi-valued image data is bit-plane encoded, m high-order bit planes and n low-order bit planes are formed at predetermined positions. (M + n = number of bit planes of multi-valued image data), the upper bit planes are encoded as they are, and the remaining lower bit planes are dithered as multi-valued image data and then encoded. A method of doing it is possible.
That is, by setting the number of bit planes to be encoded to m + 1, it is possible to reduce the data size for processing. When reproduced thereafter, the upper bit planes are decoded as they are, and the lower bit planes are reproduced by decoding the same data as n bit plane data after decoding. (That is, all the bit planes have the same data for the lower bit planes at the time of reproduction.) Therefore, in this method, only one lower bit plane is used as data, but decoding processing is performed. At this time, a memory for n bit planes is required, so there is a problem of reducing the memory required for reproduction.

Therefore, the present invention devises the processing of the bit plane data of the multi-valued image data to realize the encoding processing in which the deterioration of the image quality is suppressed and the compression rate is not lowered in each bit plane. The purpose is to further reduce the burden of processing. Another object is to reduce the amount of memory required when reproducing the multi-valued image data.

[0007]

In order to achieve the above object, the invention according to claim 1 provides an image data inputting means for inputting multi-valued image data, a binarizing means for binarizing the data,
An encoding unit for encoding and compressing the data, dividing the bit plane data of the multi-valued image data input from the image data input unit into an upper bit and a lower bit at a preset position, and the upper bit Plane data is encoded as it is by the encoding means for each bit plane, while the lower-order bit plane data is collectively binarized by the binarizing means and then encoded by the encoding means. A code processing device for multi-valued image data, which receives value image data, decodes, decompresses and outputs the coded multi-valued image data, wherein the multi-valued image data is decoded and decompressed. And two gradation values having a gradation difference of one gradation, which is composed of a high-order bit and a binarized low-order bit after decoding by the decoding means. A data conversion unit that collects the higher or lower gradation depending on the data of the upper bit and the lower bit for each, and a bit plane memory that stores the bit plane data collected by the data conversion unit. Output means for reproducing and outputting multi-valued image data by the data in the bit plane memory.

According to the first aspect of the present invention, m pieces of high-order bit plane data are encoded as they are for each plane, while n pieces of low-order bit plane data are collectively binarized into one piece. The encoded multi-valued image data (coded data) is received from a code processing device that compresses the multi-valued image data by encoding it later. This multi-valued image data becomes m + 1 plane data when decoded at this stage, but the gradation difference for each pixel of the gradation values consisting of these upper bits and lower bits when decoded is the value of the lower bits. Can be summarized because any of the binarized data differs by only one gradation. Therefore, the bit plane data of the m + 1 high-order bits and the low-order bits after decoding are collected in one of the higher and lower gray levels, and the m number of plane data are stored in the bit plane memory. And then reproduced and output. Therefore, when received, even m + 1 bit plane data is processed after the data amount is reduced to m bit planes.

According to a second aspect of the present invention, there is provided image data input means for inputting multivalued image data, binary conversion means for binarizing the data, and encoding means for encoding and compressing the data. , The bit plane data of the multi-valued image data input from the image data input means is divided into a high-order bit and a low-order bit at a preset position, and the high-order bit plane data is divided by the encoding means for each bit plane. While encoding is performed as it is, plane data of lower bits is collectively binarized by a binarizing unit, and then coded by the encoding unit and compressed. Two gradation values, which are composed of the lower bits binarized by the binarizing means and have a gradation difference of one gradation, are provided for each pixel according to the data of the upper bit and the lower bit. Provided data converting means summarized in either higher or lower of tone coding means is characterized in that the encoding bit plane data gathered by the data conversion means.

According to the second aspect of the present invention, the m high-order bit plane data of the input multi-valued image data is coded as it is for each plane, and the n low-order bit plane data are collectively collected. The data is binarized to be one and then encoded, but at this time, since the plane data of the lower bits is binarized, the grayscale value of each of the grayscale values of these upper bits and the lower bits is calculated for each pixel. The difference can be summarized because the value of the lower bit differs only in one gradation in any of the binarized data. Therefore, m +
One piece of upper-order bit plane data and lower-order bit bit plane data is grouped into one having a higher gradation or one having a lower gradation to be m pieces of plane data, and then encoded. Therefore, m + n bit plane data is
The amount of data is reduced to m bit planes.

According to a third aspect of the present invention, in addition to the configuration of the first or second aspect of the invention, the data conversion means is
The value of the lower bit binarized by the binarizing means is added to the upper bit of each pixel of the bit plane data, and when the upper bit is the maximum value of the gradation value, it is left as it is. It is characterized in that two gradation values each consisting of a bit and a binarized lower bit and having a gradation difference of one gradation are combined into a higher gradation value to determine the gradation for each pixel. To do.

According to the third aspect of the invention, in the m plane data of the upper bits and the binarized one lower bit plane data, the value of the lower bit is added to the upper bit for each pixel. When the upper bit has the maximum gradation value, it is left as it is and is grouped in the higher gradation. For example, "1" of the bit indicates high density in the image and "0" indicates low density in the image. In this case, when all the upper bits are “1”, they are left as they are, and the gradation value of the multi-valued image data consisting of m bit plane data is determined. Therefore, the invention according to claim 1 or 2 can be implemented by one adder.

According to a fourth aspect of the present invention, in addition to the configuration of the first or second aspect of the present invention, the data conversion means includes:
A lookup table that stores in advance a grayscale value associated with bit string data composed of upper bits and lower bits is provided, and the grayscale difference consists of the upper bits and the binarized lower bits. When collecting one bit plane data into either the higher gradation or the lower gradation, the lookup table is referred to according to the data of the upper bit and the lower bit, and It is characterized in that the gradation value of is determined.

According to the fourth aspect of the present invention, the look-up table is referred to for the bit string data consisting of the plane data of m high-order bits and one binarized low-order bit, and m + 1 bit strings thereof are referred to. The gradation value of the multi-valued image data is determined by the gradation value of the m bit plane data associated with the data. Therefore, it is possible to determine the gradation of the multivalued image data from the plane data of m high-order bits and 1 low-order bit.

Here, the code processing device according to the present invention means either a device for encoding multi-valued image data or a device for decoding encoded multi-valued image data.

[0016]

DETAILED DESCRIPTION OF THE INVENTION The present invention will be described below with reference to the drawings. 1 to 5 are views showing a first embodiment of a multivalued image data code processing apparatus according to the present invention.
3 corresponds to the invention described in 3. In the present embodiment, a multi-gradation image is acquired as multi-valued image data of 8 bits / pixel, and among the bit plane data of the multi-valued image data, upper 3 bits are upper bits and lower 5 bits are lower bits. The case of handling encoding / decoding will be described as an example.

First, a code processing device as an encoder having a function of encoding image data will be described. The encoder used in the description of the present embodiment is the one described in the above-mentioned problem to be solved by the invention in which the number of m + n bit planes of multi-valued image data is m + 1. In FIG. 1, an encoder 10 includes image data input means 11 for inputting multi-valued image data, and bit plane data of the input multi-valued image data for bit planes 12a to 12h (from upper bit (MSB) to lower bit ( LSB) bit plane memory 12 for storing each area prepared for each plane data) and encoding means for encoding and compressing the data
13 and a binarizing means 14 for binarizing the data.

When a multi-valued image data of 8 bits / pixel obtained by reading a multi-tone image from a scanner or the like is input to the input means 11 of the encoder 10, the image data is stored in the bit plane memory 12. After expanding (storing) each bit plane 12a to 12h, each bit plane 12a to 12h
The plane data stored in 12h is divided into two at a preset position and subjected to coding compression processing. In the present embodiment, among the plane data of the bit planes 12a to 12h, three pieces (m) are included. The upper bits of each of the bit planes 12a to 12c are coded and compressed by the encoding means 13 as they are, and the five (n) lower bits are binarized by the binary digitizing means 14 to be grouped, and then the encoding means 13 It is designed to obtain the coded data of the multi-valued image data. Thereafter, the coded data of the multi-valued image data is transmitted to the other party via the line by the communication means when the encoder 10 is installed in, for example, a facsimile machine.

Here, the encoding means 13 and the binary conversion means 14
Since it is sufficient to perform coding compression and binary conversion processing by a known method that has been conventionally performed, a detailed description thereof will be omitted, but for example, as the binary conversion means 14, a dither method or an error diffusion method, etc. Therefore, a means for converting the multivalued image into a halftone image may be used. When the entire multi-valued image data is processed by the multi-valued dither method or the multi-valued error diffusion method, the image data becomes discrete because the effect of dithering or error diffusion appears even in the upper-bit plane data, and the compression rate is increased. Will drop. However, in the present embodiment, the multi-valued image data is divided into plane data of upper bits and lower bits, and the upper bits are coded and compressed as they are for each plane data. Since the binary data is binarized, it is possible to suppress a reduction in the compression rate of the upper-order plane data and improve the compression rate of the entire image data.

Next, a code processing device as a decoder having a function of decoding image data will be described. The decoder in the present embodiment corresponds to the invention described in claims 1 and 3. In FIG. 2, a decoder 20 is a decoding means for decoding and expanding the encoded multi-valued image data.
21, a data conversion unit 22 that collects the decoded lower-order bit binarized data and the upper-order plane data,
It is provided with a bit plane memory 23 that stores the combined plane data of the upper bits and the lower data for each of the bit planes 23a to 23c, and an output unit 24 that reproduces and outputs multi-valued image data by the plane data in the memory 23. There is.

Here, the coded data processed by the decoding means 21 is, for example, data transmitted via a line or received from the storage device.
The plane data of 3 (m) high-order bits and 5 (n) low-order bits by the encoder 10 is multilevel image data in which the number of the plane data is set to m + 1. From this, when decoding the coded data and reproducing the multi-valued image data in the same manner as described in the above-mentioned problem to be solved by the invention, as shown in FIG.
The bit plane memory 123 is prepared, and the plane data of the high-order bit is simply the three bit planes 123a to 123.
Since the lower bits following this upper bit are binary data, the same data are all written and developed in the five bit planes 123d to 123h, and then all the bit planes 123a to 123h are written as one sheet. Will be output as multi-valued image data by the output means 24. Therefore, in this method, although the lower-order bit planes 123d to 123h of data are substantially one plane, a memory capacity of five bit plane data is required for the decoding process. .

However, since the decoded data processed by the data conversion means 22 is coded by the encoder 10 by the above-mentioned algorithm, at the time of reproduction processing, as shown in FIG. For one combination of 3 bits, the lower 5 bits only develop a binarized value, and thus all take a value of "0" or "1". Therefore, there is only one gray level difference between the lower bits, and even if it is combined with the upper bits, for example, as shown in FIG.
There is only one gray level difference between "00011111" and "00100000". From this, in the case of such a bit configuration,
By grouping these gradations together, it is possible to significantly reduce the bit planes required for reproduction. In other words, a set of two gradations with no such difference is
The bit plane memory used at the time of reproduction by collecting the higher or lower gradation levels.
The capacity of can be significantly reduced.

Specifically, for example, when processing each gradation set so as to be grouped into a higher gradation, when the upper bit is "000", the binarized lower bit is "0". When the lower bit is “1”, it is set to “0000” as it is, but when the lower bit is “1”, it is not set to “0001” and is set to “0010”. When the upper bit is “001” and the lower bit is “0”, "0010" is set to "0100" when the lower bit is "1", and the same applies to other combinations of upper bits, but only when the upper bit is "111" and the lower bit is "1". Since there is no gradation above it, "11
10 ". By doing this, the top 4
Since the bit position (4th digit) is always "0", only the upper 3 bits can actually be used as a gradation, and therefore the bit plane memory 23 required for reproduction is a bit from 8 planes to 3 planes. Planes 23a-23c
Can be reduced to

Therefore, the data conversion means 22 of the present embodiment
Is a combination of the upper-bit data and the lower-bit bit plane data decoded by the decoding means 21 and outputs two bit plane data whose tone difference for each pixel is one tone according to each data (state value). As shown in FIG. 3, the data conversion unit 22 binarizes the lower-order bit binarized into the upper-bit plane data from the decoding unit 21, as shown in FIG. Digitized data to adder 25
When the upper bit is "111" which is the maximum gradation value at this time, the gradation value is determined as it is,
The bit planes 23a to 23c of the bit plane memory 23 are converted so that they can be expanded. It is needless to say that the data conversion means 22 may be similarly configured so that the upper-order bit data and the lower-order bit plane data are put together in the lower gradation.

Therefore, in the decoder 20, after the code data of the multi-valued image data encoded by the encoder 10 is decoded by the decoding means 21, the data conversion means 22 converts the lower bit into the upper bit plane data. The multi-valued image data is added to the three bit planes 23a of the bit plane memory 23 by adding the binarized data and putting them together in the higher gradation.
~ 23c to output a sufficient amount of data that can be stored
It can be reproduced and output by 24. Note that, after this, the reproduced multi-valued image data is displayed by the decoder 20, for example,
In the case of being installed in a facsimile machine, the recording means records and outputs it to form a multi-tone image on a sheet.

Thus, in this embodiment, the encoder
In multi-valued image data that has been encoded and compressed by 10, the upper bit plane data is encoded as it is for each plane, while the lower bit plane data is collectively binarized and then encoded. Therefore, the gradation of each pixel obtained by decoding and combining the upper bit and the lower bit is different by only one gradation regardless of the value of the lower bit. Therefore, the data conversion unit 22 converts the binarized data of the lower bit. After being added to the plane data of the high-order bits and put together in the higher gradation, the bit plane memory 23 is expanded, reproduced and outputted by the output means 24. Therefore, the multi-valued image data encoded by the encoder 10 so as to suppress the deterioration of the image quality and prevent the compression rate from decreasing is further reduced to the plane data amount (gradation value) of the upper bits, and then the bit plane memory having a small memory capacity. The data can be expanded to 23 and played back, and the memory amount of the bit plane memory 23 can be reduced.

Further, the data converting means 22 adds the binarized data of the lower bit to the plane data of the upper bit and leaves it as it is when the upper bit has the maximum gradation value of "111". Therefore, the processing of the present embodiment can be easily performed by one adder 25. Next, FIGS. 6 and 7 are diagrams showing a second embodiment of the multivalued image data code processing apparatus according to the present invention, and the present embodiment corresponds to the invention described in claims 2 and 3. In the present embodiment, a case where the multi-valued image data of 8 bits / pixel is similarly coded and decoded by the encoder and the decoder configured substantially the same as those described in the first embodiment will be described. Since the description will be given as an example, the same reference numerals will be given to the similar configurations to describe the characteristic portions.

First, a code processing device as an encoder having a function of encoding image data will be described. In both figures, the encoder 30 includes, in addition to the image data input means 11, the bit plane memory 12, the encoding means 13 and the binarization means 14, the data conversion means 22 included in the decoder 20 in the above embodiment. Is provided.

In the encoder 30, the plane data of the five (n) lower bits of the bit planes 12d to 12h are binarized before the encoding compression processing by the encoding means 13.
After the binarization, the data conversion means 22
After being added to the three (m) upper bits 12a to 12c (when the upper bit is "111" which is the maximum gradation value, the gradation value remains as it is), it is sent to the encoding means 13. Then, the encoding / compression process is performed to convert the multivalued image data into encoded data.

Therefore, in the present embodiment, the decoder 40 is only provided with the decoding means 21, the bit plane memory 23 and the output means 24, and the data conversion means 22 is omitted. However, the code data decoded and reproduced by the decoder 40 is coded and compressed after being grouped into three bit plane data (upper bits) by the encoder 30, and thus is decoded by the decoding means 21. Alone
Three bit planes 23a of the bit plane memory 23
23c, and similarly to the above-described embodiment, the reproduction process of the code data to the multi-valued image data can be performed and output by the output means 24 with a small memory capacity.

As described above, in this embodiment, it is possible to obtain the same effects as the first embodiment, and the encoder
Since 30 allows decoding and reproduction processing with a small memory capacity at the stage of encoding and compressing multi-valued image data by the data conversion means 22, the decoder 40 undergoes special conversion processing by the data conversion means 22. It is possible to output the multi-valued image data by subjecting the received coded data to the decoding and reproduction processing as it is without being processed.

As another aspect of the present embodiment, as shown in FIG. 8, the data converting means outputs a 4-bit (digit) gradation value obtained by combining the upper bit and the binarized lower bit. Address data (bit string data) is stored in advance by associating with each other a 3-bit gradation value obtained by combining the upper bits and the binarized lower bits into the higher gradation for each address. By providing a look-up table, the look-up table is referred to according to the data obtained by combining the high-order bit and the low-order bit when the high-order bits and the low-order bits are arranged in the higher gradation, and the reproduction processing is performed. The value image data may be converted (determined) into a small number of gradation values. With such a configuration, the data conversion means can perform the same processing even when the upper bit is “111” which is the maximum value of the gradation value, and it can be processed according to the data in which the upper bit and the lower bit are combined. Since it only refers to the lookup table, it can be processed quickly. In addition, also in this other mode, the data converting means may be similarly configured so that the plane data of the upper bits and the lower bits are put together in the lower gradation. Further, it goes without saying that this other aspect may be applied to the above-described first embodiment.

[0033]

According to the first aspect of the present invention, the plane data of the high-order bits is directly encoded for each plane, and the plane data of the low-order bits is collectively binarized and encoded. (Data) is received, even if it is a gradation value that is combined after decoding the plane data of these upper bits and lower bits before outputting, only one gradation is possible because the lower bits are binarized. Since they are not different, they are grouped into one with higher or lower gradation, so that multi-valued image data encoded so that the compression rate does not decrease in each bit plane while suppressing deterioration in image quality It is possible to reduce the data amount and store the plane data in the bit plane memory for reproduction processing. Therefore, the memory amount of the bit plane memory can be reduced.

According to the second aspect of the invention, the upper-bit plane data of the multi-valued image data is encoded as it is for each plane, and the lower-bit plane data is collectively binarized and then encoded. , The gradation difference for each pixel of the gradation value consisting of the high-order bit and the low-order bit differs only in one gradation in the lower-order bit value in any of the binarized data. Since the data is grouped into one and then coded, the multi-valued image data can be reduced to a data amount of upper bits, and the multi-valued image data can be coded so that the compression rate does not decrease while suppressing the image quality deterioration. it can. Therefore, the processing can be performed without burden, and the communication time can be shortened, for example. Further, since the encoded and compressed multi-valued image data needs only the data amount of upper bits when it is decoded, a large memory amount is not required at the time of reproduction, and the memory amount is reduced in the decoding processing device. be able to.

According to the third aspect of the present invention, the plane data of the high-order bit and the low-order bit is collected by adding the binarized value of the low-order bit to the high-order bit of each pixel to obtain multi-valued image data. Is set to a gradation in the range of the maximum value to the minimum value represented by the upper bits, so that the invention according to claim 1 or 2 can be implemented with a simple configuration having only one adder.

According to the fourth aspect of the invention, the multivalued image data is converted into the gradation value associated with the bit string data of the upper bit and the binarized lower bit by referring to the look-up table. , The gradation of the multi-valued image data can be determined from the plane data of the high-order bit and the binarized low-order bit, and can be processed quickly.

[Brief description of drawings]

FIG. 1 is a diagram illustrating a first embodiment of a multivalued image data code processing apparatus according to the present invention, and is a schematic conceptual diagram of a code processing apparatus that codes multilevel image data to be decoded by the embodiment. Is.

FIG. 2 is a schematic conceptual diagram showing the first embodiment.

FIG. 3 is a conceptual diagram showing a main part of the first embodiment.

FIG. 4 is an explanatory diagram of data processed by the first embodiment.

FIG. 5 is a schematic conceptual diagram of a code processing device adopting another system for explaining the effect of the first embodiment.

FIG. 6 is a diagram for explaining the second embodiment of the code processing device for multi-valued image data according to the present invention, and is a schematic conceptual diagram thereof.

FIG. 7 is a schematic conceptual diagram showing a code processing device for decoding code data according to the second embodiment.

FIG. 8 is a conceptual diagram showing a main part of another aspect of the second embodiment.

[Explanation of symbols]

10, 30 encoder (code processing device) 11 Image data input means 12, 23-bit plane memory 13 Encoding means 14 Binarization means 20, 40 decoder (code processing device) 21 Decryption means 22 Data conversion means 24 Output means 25 adder

Claims (4)

(57) [Claims] 1. An image data input means for inputting multi-valued image data, a binarization means for binarizing the data, and an encoding means for encoding and compressing the data. The bit plane data of the multi-valued image data is divided into upper bits and lower bits at a preset position, and the upper bit plane data is directly encoded by the encoding means for each bit plane, The low-order bit plane data is collectively binarized by the binarizing means, and then encoded by the encoding means. Multivalued image data is received from the code processing device, and the encoded and compressed multivalued image data is decoded. A decoding device for decoding and expanding encoded multi-valued image data, and a decoding device for decoding the encoded multi-valued image data After the conversion, two gradation values having a gradation difference of one gradation and consisting of an upper bit and a binarized lower bit are converted into a higher gradation according to the data of the upper bit and the lower bit for each pixel. Alternatively, a data conversion unit that collects the data into one of the lower ones, a bit plane memory that stores the bit plane data collected by the data conversion unit, and an output that reproduces and outputs multi-valued image data by the data in the bit plane memory A code processing device for multivalued image data, comprising: 2. An image data input means for inputting multi-valued image data, a binarization means for binarizing the data, and an encoding means for encoding and compressing the data. The bit plane data of the multi-valued image data is divided into upper bits and lower bits at a preset position, and the upper bit plane data is directly encoded by the encoding means for each bit plane, This is a code processing device for multi-valued image data in which plane data of lower bits is collectively binarized by a binarizing means, and then encoded by the encoding means and compressed, and binarized by the upper bit and the binarizing means. The two gradation values, which are composed of the lower bits and have a gradation difference of one gradation, have a higher or lower gradation value for each pixel depending on the data of the upper bits and the lower bits. A code processing device for multi-valued image data, characterized in that data conversion means for collecting data in any one of the two is provided, and the coding means codes the bit plane data collected by the data conversion means. 3. The data converting means adds the lower bit value binarized by the binarizing means to the upper bit of each pixel of the bit plane data, and at this time, the upper bit is a gradation value. By keeping the maximum value, the two gradation values consisting of the high-order bit and the binarized low-order bit and having a gradation difference of one gradation are combined into the higher gradation, and the gradation is increased for each pixel. The code processing device for multi-valued image data according to claim 1 or 2, wherein the key is determined. 4. The data conversion means comprises a look-up table that stores in advance a gradation value associated with bit string data consisting of upper bits and lower bits, and comprises a higher bit and a binarized lower bit. When two bit plane data having a gradation difference of one gradation are put together into one of the higher gradation and the lower gradation, a look-up table according to the data of the upper bit and the lower bit. 3. The code processing device for multivalued image data according to claim 1, wherein the gradation value for each pixel of the multivalued image data is determined with reference to.