JP2003046784A - Image encoding apparatus, image encoding method, program and storage medium recording program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To carry out encoding with high-speed and high compression rate by forming the units using a plurality of pixels belonging to within one period for digital image data having a periodic property and cutting out the units, so that the same values can continue using the units as units of encoding. SOLUTION: A unit forming means 11 forms the units using the plurality of pixels out of the pixels belonging to within one period with respect to image data having a periodic property such as dither image, etc., to use the units as units of encoding. In forming the units, a statistic section 10 forms the units on the basis of a occurrence probability or a change probability indicating the possibility of changing of the encoded units into black pixels or while pixels, and adaptively determines the encoding sequence of the units to be consecutively encoded next from among the units separated by an integer multiple of the period, and cuts out the units on the basis of the determined sequence of the units to output the units to a encoding means 12. The section 12 carries out variable length encoding for the units outputted from the means 11.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image coding device,
Image encoding method, program and recording medium recording the program, and more specifically, for example, image encoding device and image encoding method for encoding an image having periodicity such as a dither image using a dither matrix The present invention relates to a program and a computer-readable recording medium recording the program.

[0002]

2. Description of the Related Art The systematic dither method is often used when outputting image data to an image processing apparatus which has a small number of expressible gradations per pixel. For example, a printer or a facsimile that prints with or without dots can usually express only two gradations per pixel. When expressing a halftone of a photograph or the like by such an image processing apparatus, an image created by the systematic dither method as shown in FIG. 20 is used.

FIG. 20 is a diagram showing an example of a dither process using the systematic dither method. 20A shows an example of multi-tone data before dither processing, and FIG. 20B shows an example of a dither matrix applied to the multi-tone data shown in FIG. 20A. 20C is a diagram showing an example of image data after the dither matrix shown in FIG. 20B is applied to the multi-tone data shown in FIG. 20A. Here, the dither processing is to compare the original multi-gradation data shown in FIG. 20 (A) and the value (threshold value) of the dither matrix shown in FIG. 20 (B) to obtain the comparison result. ,
This is a method of outputting as shown in FIG. Since the dither matrix is used repeatedly, the image created by this method has features with the same period as the dither matrix. In particular, in images such as photographs, which have a lot of low-frequency components, this feature becomes prominent.

As a method for compression-encoding a dither image having such a periodicity, there is a technique that pays attention to a dither matrix that creates the dither image. For example, JP-A-5
In the invention described in Japanese Patent No. 5-0507776, the pixels having the same phase in the dither matrix are collected and then encoded. Pixels of the same phase that are close to each other tend to have the same value, which is because the image data usually contains a lot of low frequency components. Here, if the original multi-gradation data is the same for all pixels (only the DC component is present), pixels of the same phase in the dither matrix will always have the same value.
From such a general tendency, a high compression rate can be obtained by performing run-length coding, for example, after collecting pixels of the same phase of the dither matrix.

[0005]

However, the invention described in the above publication has a problem that the processing speed is slow because the same phase is collected for each pixel.
For example, when 16 pixels of the same phase are collected and encoded, 1
If only one pixel of a certain phase can be obtained by the memory access of 16 times, the memory access of 16 times is required. Therefore, in one memory access time, only 1
Only pixels can be encoded. In order to speed up this, for example, a register for holding all phase data acquired in 16 memory accesses is prepared, or 16 pixels of the same phase are acquired in one memory access. The circuit scale had to be increased.

The present invention has been made in view of the above points, and it is possible to obtain a high speed and high compression rate with a small circuit scale for an image having periodicity such as a dither image using a dither matrix. The present invention has been made for the purpose of providing an image encoding device capable of image encoding, an image encoding method, a program, and a recording medium recording the program.

Further, according to the present invention, high-speed processing can be achieved by encoding in units of a plurality of pixels.
Furthermore, by utilizing the periodicity of the image when creating the coding unit, the types of occurrence of coding elements are aggregated to improve compression efficiency,

The compression efficiency is improved by reducing the size of a unit which is an encoding unit to 1 / M times the period to aggregate the generated patterns, and the total number of symbols generated by increasing the size to M times the period. To improve compression efficiency by suppressing

An arbitrary cycle is generated in the image data whose resolution has been converted due to the scaling. By determining the coding order of the symbols by using both the cycle generated by this resolution conversion and the cycle generated by the gradation processing after scaling, it is possible to further improve the continuity of similar patterns and perform compression. Improving efficiency,

The least common multiple of the cycle generated by the resolution conversion and the cycle of the image after gradation processing is set as the unit cutout cycle that determines the coding order, so that the continuity of similar patterns is surely improved. To further improve compression efficiency,

Depending on the value of the least common multiple, the pixels distant from each other in the main scanning direction are made to be continuous, the similarity as an actual image is lost, and the compression efficiency may be reduced. Therefore, it is necessary to set a predetermined threshold value so as to avoid a decrease in compression efficiency,

By performing resolution conversion so that the cycle of the image after gradation processing is equal to the cycle by resolution conversion, the continuity of similar patterns in the coding order is improved and the compression efficiency is improved. thing,

When changing the resolution conversion method, it is possible to prevent deterioration of the image quality of the scaled image itself by determining whether or not to change the method by setting a predetermined condition.

It is an object of the present invention to further improve the continuity of similar patterns and improve the compression efficiency by matching the cycle of resolution conversion with the cycle of the image after gradation processing.

[0015]

According to a first aspect of the present invention, a unit is configured by using a plurality of pixels belonging to one period of digital image data having periodicity, and the configured unit is used as an encoding unit. The two consecutively encoded units are separated by an integral multiple of the period.

According to a second aspect of the present invention, in the first aspect of the invention, the unit that is an integral multiple of the period and is to be encoded next is adaptively determined based on the encoded unit in the digital image data. It is characterized by that.

According to a third aspect of the present invention, in the first aspect of the invention, the pixels belonging to the unit have similar occurrence probabilities indicating the likelihood of becoming a black pixel or a white pixel. It is a thing.

According to a fourth aspect of the present invention, in the first aspect of the present invention, the pixels belonging to the unit have similar pixel value change probabilities between two pixels separated by an integral multiple of the period. It is a feature.

According to a fifth aspect of the invention, in the first aspect of the invention, the number of pixels belonging to the unit is not the same in the digital image data.

According to a sixth aspect of the present invention, in the third aspect of the invention, the occurrence probability is obtained by measuring the number of black pixels or white pixels for each phase in a cycle. It is a thing.

According to a seventh aspect of the present invention, in the fourth aspect of the present invention, the change probability is the number of pixels in which the pixel value of the next pixel, which is one cycle away, has changed for each phase in the cycle. It is characterized by what is required in.

The invention of claim 8 is characterized in that, in the invention of claim 5, the number of pixels belonging to the unit is adaptively determined based on an encoded unit in the digital image data. Is.

According to a ninth aspect of the present invention, in the first aspect of the invention, the number of pixels belonging to the unit is a divisor of the period.

The invention of claim 10 is characterized in that, in the invention of any one of claims 1 to 9, the unit is encoded.

According to an eleventh aspect of the present invention, a unit is formed by using a plurality of pixels of digital image data having periodicity, and the formed unit is used as an encoding unit. However, in the image encoding device which is N times (N is a natural number) apart from the cycle, the configuration of the unit is determined based on the cycle of the digital image data.

According to a twelfth aspect of the present invention, in the eleventh aspect, the number of pixels belonging to the unit is M times or 1 / M times (M is a natural number) the cycle of the digital image data. It is a feature.

The invention of claim 13 is the same as claim 11 or 1.
12. The invention according to claim 2, wherein the digital image data having periodicity is obtained by subjecting digital image data having a predetermined resolution to gradation processing, and the natural number N according to claim 11.
Is determined by using the image period after the gradation processing and further resolution information.

According to a fourteenth aspect of the present invention, in the thirteenth aspect, the natural number N is determined based on the least common multiple of the image period and the number of consecutive equivalence pixels corresponding to the resolution information. It is characterized by that.

According to a fifteenth aspect of the present invention, in the invention of the fourteenth aspect, the least common multiple is a threshold value Th (T
When h is a natural number), the number of pixels between two units separated by N times the period is set to a value equal to or less than the threshold Th.

[0030] The invention of claim 16 relates to claims 13 to 1.
In any one of the inventions of 5 to 5 and 5, there is a judging means for judging a predetermined condition based on the image period and the number of consecutive equivalence pixels corresponding to the resolution information, and based on the result of the judging means, It is characterized by performing the conversion of.

According to a seventeenth aspect of the invention, in the sixteenth aspect of the invention, the judging means sets, as the predetermined condition, the number of consecutive equivalence pixels corresponding to the resolution information and P times the image period, or 1. It is characterized by determining whether or not there is a natural number P whose difference from / P times (P is a natural number) is less than or equal to a predetermined threshold value Ti (Ti is a natural number).

According to an eighteenth aspect of the present invention, in the seventeenth aspect of the present invention, the predetermined conversion is such that the number of consecutive equivalence pixels corresponding to the resolution information is P times the image period or 1 / P.
It is characterized by conversion to double (P is a natural number).

According to a nineteenth aspect of the present invention, a unit is formed by using a plurality of pixels belonging to one cycle of digital image data having a periodicity, and the formed unit is used as an encoding unit, and is continuously encoded. The two units are separated from each other by an integral multiple of the period.

According to a twentieth aspect of the present invention, in the nineteenth aspect of the present invention, the unit that is an integral multiple of the period and is to be encoded next is adaptively determined based on the encoded unit in the digital image data. It is characterized by that.

According to a twenty-first aspect of the invention, in the nineteenth aspect of the invention, the pixels belonging to the units have similar occurrence probabilities indicating the likelihood of becoming a black pixel or a white pixel. It is a thing.

According to a twenty-second aspect of the present invention, in the nineteenth aspect of the present invention, the pixels belonging to the unit have similar pixel value change probabilities between two pixels separated by an integral multiple of the period. It is a feature.

According to a twenty-third aspect of the invention, in the nineteenth aspect of the invention, the number of pixels belonging to the unit is not the same in the digital image data.

The invention of claim 24 is the invention of claim 21, wherein the occurrence probability is a black pixel for each phase in a cycle,
Alternatively, it is characterized by being obtained by measuring the number of white pixels.

According to a twenty-fifth aspect of the present invention, in the twenty-third aspect, the change probability is to measure the number of pixels in which the pixel value of the next pixel, which is one cycle away, has changed for each phase in the cycle. It is characterized by what is required in.

According to a twenty-sixth aspect of the invention, in the twenty-third aspect of the invention, the number of pixels belonging to the unit is adaptively determined based on the encoded unit in the digital image data. Is.

A twenty-seventh aspect of the invention is characterized in that, in the nineteenth aspect of the invention, the number of pixels belonging to the unit is a divisor of the period.

The invention of claim 28 is based on claim 19 or 2.
The invention of any one of 7 is characterized in that the unit is encoded.

The invention of claim 29 is based on claims 1 to 18.
29. The image coding apparatus according to any one of claims 1 to 29, or the image coding method according to any one of claims 19 to 28, for realizing a function as each unit of the image coding apparatus. It is a program to do.

The invention of claim 30 is a computer-readable recording medium in which the program according to claim 29 is recorded.

[0045]

1 is a block diagram showing an example of the internal configuration of an image coding apparatus to which the present invention is applied. In FIG.
0 is a statistical means, 11 is a unitization means, and 12 is an encoding means. The statistical unit 10 extracts the characteristic information from the digital image data input as the original image data.
The unitization means 11 cuts out a plurality of pixels serving as a coding unit using the extracted feature information as a parameter. The encoding means 12 allocates an actual code for each unit composed of the plurality of pixels. Here, the digital image data is intended to have periodicity in one image data (one frame), and is, for example, an image dithered using a dither matrix or an image composed of an arbitrary repeating pattern. Any image may be used as long as it has some periodicity. Also, the invention is not limited to binary dither images. A multi-valued dither image may be used as long as it has periodicity.

In the present invention, the sequence is such that the characteristic information about the image area in which the image encoding has already been completed is used as the encoding parameter at the present time. here,
The parameters in the initial state have some specified value. According to this sequence, even if there are regions of different characteristics in one image, such as different dither matrix regions or different density regions, the characteristic information of the image region is updated at any time. It is possible to perform coding that adaptively realizes high compression by following. Of course, the present invention is not limited to this adaptive processing,
The characteristic information of the image area may be created in the first pass, and the encoding may be performed using the characteristic information in the second pass.

There is also a method in which the characteristic information about the current image area is used as the encoding parameter at the present time. In this case, the image data itself can be compressed more than the coding configuration shown in FIG. 1, but the coding parameters used must also be stored separately. Therefore, in consideration of the total compression rate, at least the coding configuration shown in FIG. 1 is superior in that the code stream can be one.

First, the unitized means 11 shown in FIG. 1 will be described as follows. The unitization means 11 includes a buffer memory for temporarily holding image data transferred in raster order. This buffer memory is, for example, input image data (1 bit / pix
The maximum period in the main scanning direction of the pixel (el) is 64 pixels (64 bits), and 1024 bits (64 pixels x 16 periods) for holding image data of at least 16 periods.
It is composed of SRAM.

FIG. 2 is a diagram showing an example of the encoding order of the unit to which the present invention is applied, and shows a state in which the binary image data of 40 pixel cycles has been transferred to the buffer memory up to 16 cycles. The unitization means 11 sequentially cuts out a plurality of pixels from this buffer memory as one unit, and outputs the cut-out unit to the coding means 12.
In the example shown in FIG. 2, eight pixels adjacent in the main scanning direction are cut out as one unit. Alphanumeric characters shown in the figure, A
1, B1, C1, ... Represent the output order of the units.

Here, in the figure, what is output next to the leftmost unit A1 is not the unit B1 on the right side thereof, but 4
The unit A2 is located 0 pixels away. The position of 40 pixels apart in the present invention means that the first pixel of the unit A1 (the leftmost pixel) is counted and the first pixel of the unit A2 (the leftmost pixel) is separated by 40 pixels. Show. This is because if the image data is created by a dither matrix having a size of 40 pixels, the image data has a period of 40 pixels, and thus the data of the unit A1 and the data of the unit A2 are likely to have the same value. . The reason is that, as described above, there are many low frequency components included in the original multi-gradation data.

If the unit data having the same value can be continuously encoded, a high compression rate can be expected. Units A1, A2,
Units B1, B2, B next to A3, ..., A16
3 ... B16, and so on. In addition, since one unit is composed of pixels whose image cycle is within one cycle, one unit can be cut out by one access from the buffer memory, which enables high-speed encoding. In this way, two consecutively encoded units are set so that they are separated by an integral multiple of the cycle of an image. As used in the present invention, the integral multiple of the period is, for example, as shown in FIG. The unit from the first pixel (the leftmost pixel) of the unit A1 to the first pixel (the leftmost pixel) of the unit to be encoded next is set as a unit, and the encoding order is set so that the units are separated by an integral multiple of the cycle. Set.

Further, by not only expecting the low frequency components of the original multi-gradation data but using a prediction method, it is possible to further increase the equivalence series of the units to be cut out. For example, the unit B2 and the unit B3 shown in FIG.
Have different black pixel numbers and are not the same value, but the unit B2 and the unit B16 have the same black pixel number and have the same value. this is,
This is because the original multi-gradation data has different densities in the areas of the unit B2 and the unit B3, but has the same density in the areas of the unit B2 and the unit B16.
Even if these two units are separated from each other by more than twice the period, if the units in the regions having similar densities are made continuous, the possibility of continuous equivalency is further increased. Here, it is possible to predict which region the density is close to from the encoded data of the adjacent phase units.

According to the example shown in FIG. 2, the units B1 ...
The cutout order of B16 is predicted from the data of the units A1 to A16. For example, there is a method of cutting out the units B1 to B16 in descending order of the number of black pixels in the units A1 to A16. In the example shown in FIG. 2, the unit A3 has two black pixels, but the other units A1, A2, and A16 have four black pixels, and if there is no other unit having four black pixels. , Unit B2 and unit B16 with three black pixels
Will be cut out continuously. According to this method, since the encoded unit data is used, it is not necessary to newly save the information indicating in what order the units B1 to B16 were cut out. This is because when decoding these units, the order of the decoded units A1 to A16 to the units B1 to B16 can be restored by using the same rule as in the encoding.

Here, the statistical means 10 shown in FIG. 1 determines the number of black pixels and the order of encoding. Instead of sorting the units based on the number of black pixels, it is determined whether or not the number of black pixels in each unit exceeds a predetermined value, first cut out units with a predetermined value or less from the left, and then exceed the predetermined value. There is also a method of cutting out units from the left. According to this method, the possibility of continuous equivalence is somewhat reduced, but complicated sort processing can be omitted. Thus, the coding order of the units can be adaptively determined by predicting from the coded area.

FIG. 3 is a diagram showing an example of pixel positions of a unit to which the present invention is applied. In this example, the configuration pattern of the units belonging to one period will be described with one period of 40 pixels. Until now, as shown in FIG. 2, the unit A1, the unit B1, the unit C1, the unit D1, and the unit E1 were cut out as shown in FIG. However, the present invention is not limited to this. As shown in FIG. 3B, it is not always necessary to form a unit with adjacent pixels.
You may comprise a unit with the pixel which does not adjoin. However, as described above, the pixels belonging to each unit are assumed to belong to one cycle of the image. Further, as shown in FIG. 3C, the unit may be composed of, for example, 16 pixels instead of 8 pixels. Here, the pixels forming the unit are selected on the basis of statistical information values as shown in FIG. 4 described later.

FIG. 4 is a diagram showing an example of processing for determining the pixel position of a unit to which the present invention is applied. In this example, the statistical information value included in a unit belonging to one period will be described assuming that one period is 40 pixels. The statistical information value given to each pixel shown in FIG. 4A has a value for each phase within one cycle of the image. Here, the statistical information value is a probability of occurrence of black pixels or white pixels, or
It is a value that is determined based on a change probability that indicates how easily the pixel value changes between adjacent cycles. In the example shown in FIG. 4A, one period is 40 pixels, and 0 to each pixel in one period.
Up to 16 statistical information values are assigned. This statistical information value is a value based on the occurrence probability and shows a tendency of black pixels or white pixels to be easily formed. It is considered that the larger the number is, the more easily black pixels are formed. By doing so, it is possible to increase the possibility of continuous equivalence.

FIG. 4B shows the result of rearranging the statistical information values shown in FIG. 4A in order of magnitude. In this way, the one-to-one rearrangement is performed on the actual image data as shown in FIGS. 4 (C) and 4 (D). FIG. 4C shows an arrangement example of black pixels and white pixels in actual image data of 40 pixels per cycle, and in the figure, gray pixels are black pixels. FIG. 4D is a rearrangement of the actual image data shown in FIG. 4C based on the occurrence probability, and white pixels and black pixels can be made continuous. However,
One white pixel is arranged in the continuous black pixels shown in (D), but this is considered to be a pixel that statistically causes an error.

In the state shown in FIG. 4D, the adjacent pixels are solidified to form a unit. 1 due to a pixel with a phase that has a large deviation in the occurrence probability such that it tends to become a black pixel or a white pixel, or a pixel with a phase with a low change probability such that the value does not change much between adjacent periods If a unit can be configured, the unit is likely to have the same value. As compared with the case where adjacent pixels in the image space are used as a unit as shown in FIG. 3A, the complexity is slightly increased, but a higher compression rate can be expected. As described above, by selecting the pixels forming the unit based on the occurrence probability or the change probability, it is possible to increase the possibility of continuous equivalence and improve the compression rate.

If the above-mentioned statistical information value is used as known fixed data and is commonly used for different types of images, the above-described effect of improving the compression ratio may not be obtained depending on the image. Also, even if the fixed data is different for each image, if there are areas of different properties mixed in the image,
The same problem occurs. Therefore, if the statistical information value is adaptively created from the local area of the image itself that is the compression target, this problem does not occur, and a stable compression rate improvement effect can always be obtained.

FIG. 5 is a diagram showing an example of measurement of statistical information values to which the present invention is applied. An example of measuring the statistical information value by obtaining the occurrence probability and the change probability from the adjacent coded image data will be described. In the example shown in FIG. 5, a statistical information value is measured from 16 pixels at intervals of 40 pixels, using an area for 16 cycles that has been encoded (a region for 640 pixels) that is close to image data of 40 pixels as a learning area. In the figure, the vertical axis represents the phase of pixels within one cycle (40 pixels in this example),
The horizontal axis represents a cycle of 1 to 16 cycles. FIG. 5 (A),
Although four phases are shown as shown in (B), (C), and (D), the same measurement is actually performed for one cycle, that is, for all 40 phases. First, FIG.
The phase shown in (A) is a white pixel over all 16 cycles, and it is always judged that the same value can be expected. Figure 5
When (B) and the phases shown in FIG. 5 (C) are compared, it is determined that the occurrence probabilities are not biased, so that it is impossible to expect the same value to continue in both phases.
It is judged that the phases shown in (B) can be expected to have the same value.

However, if the phase shown in FIG. 5 (B) and the phase shown in FIG. 5 (D) are the same unit, they do not become the same value continuous, so the phase shown in FIG. 5 (B) becomes the same value continuous. You can't judge that you can expect.
From these things, the AND condition of the occurrence probability and the change probability is considered to be the most correct judgment. However, considering the effect obtained, it is considered sufficient to measure only one of the two probabilities. The statistical means 1 shown in FIG. 1 is used to measure these statistical information values and determine the unit configuration based on the measurement results.
Perform at 0.

Since the statistical information value is measured from the encoded learning area, the decoding can be performed without storing the result separately. On the other hand, the encoded learning area and the encoding target area that reflects the learned result in the unit configuration are close to each other but different areas, so if the image features of both areas are different, Proper unit configuration cannot be done. However, it is considered that the effect on the compression rate is small when there are few portions of the image whose characteristics change. Also,
The adaptive processing has a small effect on the processing speed. In the above-mentioned example, since the same statistical information value is used while the image data for 16 cycles input to the buffer memory is cut out, the time spent for the one-to-one rearrangement conversion accompanying the update of the statistical information value. The effect of is also minor. As described above, the probability of occurrence or the probability of change is created from the local area of the image to be encoded, and the pixels forming the unit are adaptively selected based on these to increase the possibility of continuous equivalence and improve the compression rate. be able to.

In the embodiment described so far, one unit is cut out with eight pixels fixed, but the present invention is not limited to this. As shown in FIG. 3C described above, a unit configured with a different number of pixels may exist in one image cycle. The number of pixels forming each unit is, for example,
The statistical information value shown in FIG. 4A is selected as a reference. Different encoding is performed depending on whether or not this value exceeds a predetermined value. In the example shown in FIG. 3 (C), 16 pixels which can be expected to have the same value continuity of a predetermined value or less are set as one unit, and other units are set as 8 pixels. Since the portion where this one unit has 16 pixels performs different encoding from the remaining portion, the circuit size does not increase even if one unit has 16 pixels. In this case, as compared with the unit configuration shown in FIG. 3A, 16 pixels can be encoded at one time, so that the processing can be speeded up. In this way, the processing speed can be increased by limiting the number of pixels forming the unit and allowing the unit including a plurality of different numbers of pixels to exist.

Rather than using the above-mentioned statistical information values as known fixed data, it is always possible to obtain a stable compression rate improvement effect by adaptively creating from the local area of the image itself to be compressed. It is clear from the above description.
The statistical information value is measured based on the occurrence probability or the change probability from the adjacent coded image data, and the unit is selected based on this statistical information value. In this way, the occurrence probability or change probability is created from the local area of the encoding target image,
Based on this, the number of pixels forming the unit is adaptively changed.

Further, it is also possible to use pixels within a cycle as the pixels constituting the unit and make the number of pixels a divisor of the cycle. For example, if the period is 40 pixels, 2, 4, 8, 2
A unit is composed of 0 pixels. This is because the image data is repeated based on the cycle, and by separating the units for each divisor of the cycle, the correlation between the units can be increased and the compression efficiency can be expected to improve.

Here, the encoding means 1 shown in FIG.
2 will be described below. In this example, MTF (Mov
An example of performing Golomb-Rice encoding after e-To-Front) processing and run-length processing will be described. First, the MTF process is a process of converting the unit data cut out by the unitizing means 11 into an index of a dictionary in which symbols matching the unit data are stored in advance. For example, when the input unit data is 8 bits, the dictionary has 256 indexes, and symbols of all 256 types of units are stored here. When the input unit data matches the dictionary data of the index N, the dictionary data of the index N is output and at the same time, the matched dictionary data is moved to the index 0,
.. move the dictionary data of index N-1 one by one toward a larger number. As the processing progresses, frequently-appearing dictionary data gather at the positions of small index numbers, so that input data having an arbitrary distribution shape can be converted into output data having a shape in which 0 is gradually reduced to a peak.

Further, the MTF process can be said to be a process for outputting how many kinds of input symbols the current input symbol matches, and is a BSTW (Bentley-Sleat).
or-Tarjan-Wei) code, Recency
Also called Rank code.

Next, run-length processing is performed on the output data after MTF processing. The smaller the value of the MTF-processed output data is, the higher the occurrence probability is. Therefore, the smaller the value, the higher the possibility that the same value continues. Here, 0
An example is shown in which the run count is performed only for a series of. For example 0,
4 events called 0, 0, 0 are converted into 1 event called run4. Furthermore, event integration of run and index is also performed for the Golomb-Rice encoder described later. For example, input of 0, 0, 0, 0, 3, 0, 0, 1, 5, 0, 2 is (run4, index3), (run2, in
dex1), (run0, index5), (run
1, index 2).

FIG. 6 is a diagram showing an example of the order of event occurrence probabilities. In the figure, the vertical axis is index and the horizontal axis is runlength.
As h, the intersection is the input data n of the code table shown in FIG.
And FIG. 7 is a diagram showing an example of the Golomb-Rice code table. A unique code is assigned to each event. Here, in the Golomb-Rice code shown in FIG. 7, different codes can be created with a parameter of order, so that if the parameter suitable for the distribution shape of the occurrence probability of the input data n is selected, the coding efficiency is improved. You can Further, FIG. 6 shows the relationship between the event (run_, index_) and the input data n of the code table shown in FIG. 7. All the combinations of (run_, index_) are numbered in the order of their occurrence probabilities and used as input data n. Actually run = 0 ~ 15, index
= 1 to 255, the input data n is 0 to 407.
However, in the example shown in FIG. 6, run = for the sake of explanation.
0 to 7 and index = 1 to 8 are set. Here, the coding means used in the present invention is not limited to this example, and other coding means can be used, and the compression rate can be improved by applying various coding means alone or in combination. You can

FIG. 8 is a diagram showing an example of an image cycle detection circuit. In the above description, an example in which the cycle of an image such as 40 pixels is known has been shown.
This cycle may also be adaptively detected from the image to be encoded by a circuit as shown in FIG. In the example shown in FIG. 8, only four typical image periods are detected,
The XNOR operation circuit between two pixels, which are separated from each other by the number of pixels representing the image cycle, is moved by one pixel on the image for a predetermined distance, and the cycle in which the value of the counter for adding the operation result is the largest is set as the cycle. Further, although the one-dimensional example is shown as the image cycle, the present invention is not limited to this. The unit may be cut out from a rectangular area of one cycle in the horizontal direction / a plurality of pixels in the vertical direction. Further, the target image data is not limited to the binary dither image shown here, and the same encoding can be performed on a multi-level dither image as long as the image has some periodicity. Here, a dithered result of 0 or 1, that is, 1 bit is set as a binary dither, and an Nbit (2 ≦ N) is set as a multilevel dither.

FIG. 9 is a block diagram showing an example of the internal structure of an image coding apparatus according to another embodiment of the present invention.
Reference numeral 0 is a period detecting means, 21 is a unitizing means, 22 is a unit cutting-out means, and 23 is an encoding means. The input original image data is image data which is composed of a plurality of bits per pixel and has some periodicity. For example, it is assumed that gradation processing such as dither is performed, or that the data series is in a predetermined format by block coding or the like. Here, in FIG.
The following description will be made on the basis of 1-bit / pixel data having periodicity by the systematic dither method as shown in FIG.

In the image data, the unitizing means 21 groups a plurality of pixels into one unit, and this unit is used as an encoding unit. The number of pixels forming the unit is determined based on the cycle of the input image data. Since the matrix size of the dither used is known, this cycle may be given to the unitization means 21 as a parameter, or may be automatically detected by the cycle detection means 20.

FIG. 10 is a diagram showing a structural example of the cycle detecting means 20. A certain finite number of dither periods N (here, 1
6, 20, 28, and 32), and paying attention to the value of each pixel that is separated from the target pixel by each cycle, while shifting the circuit in units of one pixel as shown in FIG. Will be counted. As a result, it can be seen that the dither cycle of this image is the one with the largest count among the N kinds of dithers, that is, the one with the most stable data (= no change). By using such an automatic detection means, it is possible to deal with image data in which a plurality of dithers are mixed at any time. At this time, since the unit size changes as the cycle changes, it is necessary to add a predetermined bit indicating the change to the code. In this way, a unit is formed in the unitization means 21 that has obtained the cycle size.

FIG. 11 is a diagram for explaining an example of the cutout method for determining the unit configuration and the coding order. FIG. 11A is a diagram showing a case where an image has a small cycle size, that is, an image formed of a small number of pixels. In this case, the size (= number of pixels) of the unit that is the encoding unit is M times the cycle (M is a natural number).
It is good to This is because if one coding unit is composed of a plurality of pixels as much as possible in consideration of the speeding up of the coding process, the total number of coding targets can be reduced, leading to speeding up and also improving compression efficiency. Because. Here, based on the parameter “cycle = 4”, “M = 2”
The 8 pixels are configured as one unit. Of course, for example, “M = 3” may be multiplied, but if the natural number M is made too large, the generation types of the unit will be wide, and the coding efficiency will be lowered.

FIG. 11B is a diagram showing the case where the cycle size of the image is large, that is, the image is composed of a large number of pixels. In this case, the period is 1 / M times (M is a natural number). In the example shown in FIG. 11B, the cycle is 16
Since it is a pixel, 8 pixels of “1/2” when M = 2 are set as one unit. The natural number M is set to a number that divides the cycle. Note that, as in the case where the cycle size of the image shown in FIG. 11A is small, if the natural number M is made too large, the size of one unit becomes small,
Care must be taken as it will hinder the speedup. As a method of creating these units, M times or 1 / M times can be used depending on the processing unit. Decoding is possible by switching to a block unit of processing and adding a predetermined discrimination bit to the code. Also, in the example shown in FIG. 11, the unit size is created as “8”,
It may be "16" or "32", and may have any number of pixels.
For this value, a size that is efficient in actual processing may be appropriately selected.

When the method of creating the unit is determined as described above, the unit cutting-out means 22 cuts out the unit based on the encoding order. FIG. 11 described above.
(A) shows an example in which the unit is continuously cut out. This is called the phase number 1. In addition, FIG.
Shows an example of cutting out one by one.
This is called the phase number 2. This cutout is completed for each block of the processing unit. When skipping one unit, start from "Unit 0" and end at "Unit N-".
When "1" is reached, the next "Unit 1" that has been skipped is sequentially cut out, and the same is cut out to the last "Unit N". The block size, which is a unit of encoding processing, may be, for example, one line, one line or less, one page, or any size.

The unit is cut out in order to improve the coding efficiency, and is intended to make the same value or an approximate value continuous. That is, this object can be achieved by cutting out using the cycle of the image. The cut-out unit is encoded by the encoding means 23. Here, by using a dictionary, for example, the Move-To-Front method or the like, the compression efficiency can be further increased. This M
The ove-To-Front method converts a generated value into a dictionary index and outputs it. Then, an update type dictionary is used in which the dictionary is updated while registering the current value at the beginning of the dictionary. As a result, it is possible to convert the data to be distributed around the index 0. The data converted in this manner is further subjected to coding by Huffman coding or the like with further improved compression efficiency using, for example, the run length method or the like.

FIG. 12 is a block diagram showing an example of the internal structure of an image coding apparatus according to another embodiment of the present invention.
Reference numeral 30 is a resolution conversion means, 31 is a gradation processing means, 32 is a unitization means, 33 is a unit cutting-out means, 34 is an encoding means, 35 is a cycle detection means, and 36 is resolution conversion information creation means. The original image data in this embodiment is
It is assumed to be composed of 8 bits / pixel. The original image data can be input from various input means, for example, image data input from a digital camera or a PC.
It is assumed that the data described in PDL (Page Description Language) by the above application or the like is created at an arbitrary resolution. This original image data is output as
For example, when outputting to a printer having an arbitrary resolution, the resolution conversion information creating means 36 on the printer driver installed on the PC side converts the original image data into n pixels per pixel based on the resolutions of the input means and output means. A scaling factor of multiplying is calculated. Using this result, the resolution conversion means 30 enlarges and reduces the original image data to develop it into a bitmap. This enlargement method is enlarged using, for example, the nearest pixel replacement method.
After the scaling, the gradation processing unit 31 applies a systematic dither method or the like to binarize or N-value. The subsequent processing is the same as the processing described in FIG. 9 described above, except that the unit cutting-out means 33 determines the unit cutting-out method using the scaling factor calculated by the resolution conversion information creating means 36. That is the point.

FIG. 13 is a diagram for explaining another example of the unit cutting-out method. In this example, when the scaling factor obtained by the resolution conversion information creating means 36 is "6",
That is, the case where the information is to expand to 6 pixels per pixel will be described. In this case, the image after scaling is X
The data has 6 pixels in the direction and 6 pixels in the Y direction. That is, it can be said that the expansion cycle is "6".
The gradation processing means 31 performs dithering processing on this image, and the cycle detection means 35 automatically detects the cycle of the image data. Of course, as described above, the dither matrix size may be obtained as a known value and used as the image period. As a result, it is assumed that the image period is “4”. Here, the unit is 1 times the image period,
That is, a case where four pixels are set as one unit will be described as an example. The least common multiple of the image period and the enlargement period is "12". That is, the 12th pixel may be cut in the image cycle,
In addition, the enlargement cycle is also in a good position. By utilizing this position, two pieces are cut out as shown in FIG.
In this case, the number of phases of the unit is three. In this way, by using the expansion period as well, it is possible to connect similar units with higher accuracy than when determining the extraction method using only the period of the image, and it is possible to improve the coding efficiency.

In the present invention, the value of the least common multiple is
When it exceeds a predetermined threshold value (Th: natural number), the distance between the cut-out units can be limited. For example, when the enlargement cycle = 19, the image cycle = 20, and the unit size is 20 pixels, the least common multiple becomes a value of “380”, which means that the units at a distance of 380 pixels are continuous. . The further the distance is in the X direction, the worse the correlation of pixels tends to be. That is, instead of making similar units continuous, units having no similarity may be made continuous and the compression efficiency may be reduced. Therefore, the distance is limited and, for example, the threshold value Th is set to 100. In this case, the image cycle is "2
Since it is "0", it becomes a cutout with four skipped phase numbers. In this case, although the enlargement period cannot be used, it is considered that the disadvantage is that the enlargement period is used. Therefore, it can be said that the use of only the image period is more effective. Alternatively, in the above example, the unit size is set to 20 pixels, but it is 1/2, which is 10 pixels.
It can also be a pixel. This makes it possible to reduce the least common multiple of the enlargement ratio. The least common multiple in this case does not fall below the threshold Th, but it is effective in some cases.

FIG. 14 is a block diagram showing an internal configuration example of an image coding apparatus according to another embodiment of the present invention.
40 is a resolution converting means, 41 is a gradation processing means, 42 is a unitizing means, 43 is a unit cutting-out means, 44 is an encoding means, 45 is a judging means, 46 is image period information, 47
Is resolution conversion information. The present embodiment shows an example of the configuration when the period of an image such as the dither matrix size is known. The image cycle may be detected by using the cycle detection unit 35 shown in FIG. However, in that case, it may be necessary to redo the resolution conversion itself. However, in the case of an image in which a plurality of dithers are mixed, this is more accurate. The difference from the device configuration shown in FIG. 12 described above is that a determination unit 45 that receives two pieces of information, that is, image period information (dither matrix size, etc.) and resolution conversion information is provided, and the resolution conversion unit 4 is determined based on this result.
The conversion method at 0 is changed.

As a concrete judgment condition of the judging means 45, for example, when the image cycle is 18 pixels and the enlargement cycle is 16 pixels, the difference is 2 pixels. For this difference, a predetermined threshold value Ti (Ti is a natural number) is set, and for example, when the threshold value Ti = 5 or less, the conversion method is changed. In the case of this example, since the condition is hit, the resolution conversion method is switched, and conversion is performed using a method different from the usual method. Alternatively, when the image period is 8 pixels and the enlargement period is 18 pixels, the difference is large as it is, so it is a good idea to make the determination as "16", which is twice the image period. By providing this threshold value Ti,
It is possible to suppress the deterioration of image quality when the enlargement method is changed. However, conversely, if Ti is increased, the error also increases and the enlarged image may deteriorate.

FIG. 15 is a diagram showing an example of enlargement processing in the resolution conversion method. Specifically, since the image cycle is 16 pixels, the expansion cycle is basically 16 pixels. FIG. 15 (A) shows a normal enlargement example, and each pixel is enlarged to 18 pixels. In the figure, A, B, C, ... Show pixel values. FIG. 15B shows an enlarged example using the above conversion method. This is a method of enlarging the same original image data based on 16 pixels per pixel. 2 per pixel of original image data
Since an error for pixels occurs, it must be absorbed somewhere. At the time when the pixel value E1 shown in FIG. 15 (B) is processed, an error of 10 pixels occurs as compared with the normal processing shown in FIG. 15 (A). Conditions are provided such that the error is absorbed when the error becomes equal to or larger than a predetermined threshold value Tj (Tj is a natural number). In this example, the pixel value E1,
We are doing this at E2. Actually, the scaling factor will be different in this part, but it is necessary to set Tj in a range that does not cause a visual problem and to determine the error absorbing method. Of course, the resolution conversion method is not limited to this, and any method other than the above can be used.

FIG. 16 is a flow chart for explaining an example of the image coding process to which the present invention is applied. First, in digital image data having a cycle to be encoded, the number of pixels corresponding to the cycle of the image data is set (step S1). Here, the number of pixels to be set is, for example, 40 pixels as a cycle according to the size of the dither matrix, as shown in FIG. Next, a plurality of pixels within one cycle (40 pixels) are sequentially cut out to form a unit, and for example, 8 pixels are configured as one unit as shown in FIG. 2 (step S2). However, the number of pixels forming the unit is not limited to this, and can be set according to the characteristics of the image data. For example, when the period is 40 pixels, its divisor is 2,
One unit may be composed of 4, 8, or 20 pixels. This is because the correlation between units can be increased and the compression efficiency can be expected to be improved by setting the number of pixels forming the unit to be a divisor of the image period.

The unit to be output next to the constructed unit is not the immediately adjacent unit, but the unit separated by 40 pixels set in step S1 is output next (step S3). This is because there is a high possibility that the unit data 40 pixels apart have the same value by setting the number of pixels according to the cycle of the dither image. By doing so, since unit data having the same value can be continuously encoded, a high compression rate can be expected.

Further, in this example, the MTF (Move-To-Front) described above is used for the unit data encoding process.
After t) processing and run-length processing, Golomb-Rice coding is performed on these combinations (step S4).

FIG. 17 is a flow chart for explaining another example of the image coding process to which the present invention is applied.
The image encoding process shown in FIG. 16 described above was expected for the low frequency components of the original image data.
The image encoding process shown in (1) uses a prediction method to further increase the equivalence sequence of unit data, and adaptively determines an integer multiple of the cycle to output the next unit. Even if the unit data are distant from each other, if the units having similar densities are extracted and made continuous, the possibility of continuous equivalence is further increased.

First, the number of pixels corresponding to the cycle of the image data is set in the digital image data having the cycle to be encoded (step S11). For example, as in the example shown in FIG. 9, one period is set to 40 pixels. Then, a plurality of pixels belonging to one cycle (40 pixels) form one unit (step S12). next,
Prediction is made from which unit density is close, from the already encoded adjacent unit data. For example, the number of black pixels in the coded unit is counted, and the units are cut out in descending order of the counted number. As described above, the equivalence succession can be increased by adaptively using the prediction method and determining the unit to be output next by being an integer multiple of the cycle (step S13).

Next, the unit to be output next to the unit configured in step S12 is set to output based on the cutout order determined in step S13 (step S14). Hereinafter, with respect to the encoding means, similar to the processing shown in FIG. 9, the output unit data is subjected to, for example, MTF (Move-To-Front) processing, ru.
After the n-length processing, Golomb-Rice encoding or the like is performed (step S15). The encoding means of the present invention is not limited to this.

Here, as described above, the pixels forming the unit are not limited to 8 pixels. For example, by using the statistical information value based on the above-mentioned occurrence probability or change probability,
Pixels having similar values may be collected to form a unit. Further, the pixels forming the unit do not necessarily have to be adjacent to each other, and rearrangement can be performed based on the statistical information value to increase the possibility of continuous equivalency. However, the pixels forming each unit are assumed to belong to one cycle of the image data.

FIG. 18 is a flow chart for explaining another example of the image coding process to which the present invention is applied.
First, the cycle in the image data is detected (step S
21). For this, the dither matrix size is obtained as a known value, and this may be used as the cycle, or the cycle detection means 35 may automatically detect the cycle. After detecting the cycle of the image data, it is determined whether the number of pixels corresponding to the cycle is equal to or larger than a predetermined value (step S2).
2). If the number of pixels is equal to or larger than the predetermined value (YES), 1 / M times the cycle is set as one unit (step S23). When "8" is set as the predetermined value, for example, the image data having a period of 16 pixels is set to have, as one unit, 8 pixels which is, for example, 1/2 times the period. If the number of pixels corresponding to the cycle is less than or equal to the predetermined value (NO) in step S22,
M times the cycle is set as one unit (step S2
4). Similarly to the above example, when "8" is set as the predetermined value, the image data of 4 pixels in the cycle has, for example,
The unit is set to double 8 pixels. As for the setting of the coding unit, it is possible to set the number of pixels, which is efficient in actually performing the processing, according to the cycle.

Next, it is judged whether or not the resolution conversion information including the scaling factor when the resolution conversion is performed on the image data is used (step S25), and when the resolution conversion information is used (YES), The number of consecutive equivalence pixels that have been scaled by this resolution conversion (hereinafter referred to as the enlargement period)
And the image period generated by the dither processing or the like are used to determine the coding order. At this time, it is determined whether the least common multiple of the enlargement cycle and the image cycle is less than or equal to the threshold Th (step S26). If the least common multiple is less than or equal to the threshold Th (YES), cycle information (image The cutting order is determined based on the cycle) and the resolution conversion information (enlargement cycle) (step S28). Step S2 above
When the resolution conversion information is not used in 5 (NO), the cutout order is determined based only on the cycle information (step S27), and the process proceeds to step S30.

If the least common multiple is greater than or equal to the threshold Th (NO) in step S26, the number of pixels between units is set to Th or less (step S29), the process proceeds to step S27, and the enlarging period is expanded. Is not used, the cutting order is determined only by the cycle information of the image cycle. At this time, changing the unit size based on the image period to reduce the least common multiple with the enlargement period may be effective even if the threshold Th or more. Next, the unit is cut out based on the cutting order (step S30). Furthermore, in this example, the MTF (Move-To-Front) described above is used for the unit data encoding process.
After t) processing and run-length processing, Golomb-Rice coding is performed on these combinations (step S31).

FIG. 19 is a flow chart for explaining another example of the image coding process to which the present invention is applied.
First, the cycle in the image data is detected (step S
41). For this, the dither matrix size is obtained as a known value, and this may be used as the cycle, or the cycle detection means 35 may automatically detect the cycle. After detecting the cycle of the image data, it is determined whether the number of pixels corresponding to the cycle is equal to or more than a predetermined value (step S4).
2). If the number of pixels is equal to or larger than the predetermined value (YES), 1 / M times the cycle is set as one unit (step S43). When "8" is set as the predetermined value, for example, the image data having a period of 16 pixels is set to have, as one unit, 8 pixels which is, for example, 1/2 times the period. If the number of pixels corresponding to the cycle is less than or equal to the predetermined value in step S42 (NO),
M times the cycle is set as one unit (step S4
4). Similarly to the above example, when "8" is set as the predetermined value, the image data of 4 pixels in the cycle has, for example,
The unit is set to double 8 pixels. As for the setting of the coding unit, it is possible to set the number of pixels, which is efficient in actually performing the processing, according to the cycle.

Next, the input of the period information regarding the period of the image data and the resolution conversion information including the scaling factor at the time of performing the resolution conversion is accepted (step S45), and the equivalent pixel generated by the scaling by this resolution conversion is received. The coding order is determined by using both the number of consecutive times (hereinafter referred to as an expansion period) and the image period generated by dither processing or the like. On this occasion,
It is determined whether or not the difference between the number of pixels in the enlargement period and the number of pixels in the image period is less than or equal to the threshold value Ti (step S4
6) If the difference is less than or equal to the threshold value Ti (in the case of YES), the resolution conversion method is switched, and conversion is performed using a method different from the normal method (step S47), and cycle information (image cycle) and resolution conversion are performed. The cutout order is determined based on the information (enlargement cycle) (step S49). Step S above
At 46, if the difference is greater than or equal to the threshold value Ti (NO
In the case of), the image cycle is changed (step S48), and the process returns to step S46 to perform the processing. This is because, for example, when the image cycle is 8 pixels and the enlargement cycle is 18 pixels, the difference is large as it is, so the image cycle is doubled to “16”, for example.
This is because it is preferable to make a decision as.

Next, units are cut out based on the cutting order determined in step S49 (step S50). Further, in this example, the MTF (Move-To-Front) described above in the unit data encoding process is used.
After the processing and the run-length processing, Golomb-Rice coding is performed on these combinations (step S51).

Although the respective embodiments have been described above centering on the coding processing of the image coding apparatus according to the present invention, the present invention has been described with reference to each function of the image coding apparatus as a method. It can also take the form of an image encoding method for realizing the encoding process. Further, the present invention may be embodied as a program for causing a computer to execute the image encoding method for implementing the encoding process.

An embodiment of a recording medium storing a program for realizing the encoding process of the image encoding device according to the present invention will be described below. As the recording medium, specifically, a CD-ROM, a magneto-optical disk, a DVD-ROM,
Floppy disk, flash memory,
Memory cards, memory sticks and other various ROMs
A RAM or the like can be assumed, and a program for causing a computer to execute the functions of the above-described embodiments of the present invention on these recording media, recording a program for realizing the encoding processing of the image encoding device, and distributing the program. , Facilitate realization of the function. Then, the recording medium as described above is attached to an information processing device such as a computer and the program is read by the information processing device, or the program is stored in a storage medium such as an HDD included in the information processing device. By reading out as needed, the encoding process according to the present invention can be executed.

[0099]

The effects of the present invention are set forth in claims 1, 10, 19, 28, 29 and 3.
According to the invention of No. 0, the periodicity of the image data is used to perform encoding with a plurality of pixels as a single unit, so that encoding with a high compression rate can be realized with a high-speed and small-scale circuit. it can.

According to the inventions of claims 2 and 20, since the coding order of pixels is adaptively changed by prediction, the coding can be realized at a higher compression rate.

According to the inventions of claims 3, 4, 21 and 22, since the pixels which are likely to have the same value continuity are used as the coding unit, the coding can be realized at a higher compression rate.

According to the inventions of claims 5 and 23, since the coding unit is set according to the number of pixels which are likely to have the same value continuity, higher speed coding can be realized.

According to the inventions of claims 6, 7, 8, 24, 25, and 26, since the coding unit is adaptively changed, the coding can be realized at a higher compression rate.

According to the ninth and the twenty-seventh aspects of the present invention, the number of pixels forming a unit is set to be a divisor of the cycle of the image, whereby the correlation between the units can be increased and the compression efficiency can be improved.

According to the eleventh aspect of the present invention, it is possible to perform high-speed processing by encoding in units of a plurality of pixels, and further by utilizing the periodicity of the image in creating the encoding unit, the encoding element It is possible to improve the compression efficiency by aggregating the types of occurrence of.

According to the twelfth aspect of the present invention, the compression efficiency is improved by reducing the size of the unit which is the coding unit to 1 / M times the cycle and consolidating the generated patterns, and to the M times the cycle. The compression efficiency can be improved by increasing the number of total symbols generated.

According to the thirteenth aspect of the present invention, the coding order of the symbols is determined by utilizing both the cycle generated by the resolution conversion and the cycle generated by the gradation processing after scaling. It is possible to improve continuity of similar patterns and improve compression efficiency.

According to the fourteenth aspect of the present invention, the least common multiple of the period generated by the resolution conversion and the period of the image after gradation processing is set as the unit cutout period for determining the encoding order, and thus, the similarity is obtained. It is possible to surely improve the continuity of the pattern and further improve the compression efficiency.

According to the fifteenth aspect of the present invention, depending on the value of the least common multiple, pixels which are distant in the main scanning direction are made to be continuous with each other, the similarity as an actual image is lost, and the compression efficiency is reduced. Will decrease
By providing a predetermined threshold value, it is possible to avoid a decrease in compression efficiency.

According to the sixteenth aspect of the invention, the resolution conversion is performed such that the cycle of the image after gradation processing is equal to the cycle of the resolution conversion, thereby improving the continuity of similar patterns in the encoding order. Thus, the compression efficiency can be improved.

According to the seventeenth aspect of the present invention, when the resolution conversion method is changed, it is determined whether or not the method is changed by setting a predetermined condition to prevent deterioration of the image quality of the scaled image itself. be able to.

According to the eighteenth aspect of the present invention, by making the cycle of resolution conversion and the cycle of the image after gradation processing coincident, the continuity of similar patterns can be further improved and the compression efficiency can be improved. You can

[Brief description of drawings]

FIG. 1 is a block diagram showing an example of the internal configuration of an image encoding device to which the present invention is applied.

FIG. 2 is a diagram showing an example of an encoding order of units to which the present invention is applied.

FIG. 3 is a diagram showing an example of pixel positions of a unit to which the present invention is applied.

FIG. 4 is a diagram showing an example of processing for determining a pixel position of a unit to which the present invention is applied.

FIG. 5 is a diagram showing an example of measurement of statistical information values to which the present invention is applied.

FIG. 6 is a diagram showing an example of the order of event occurrence probabilities.

FIG. 7 is a diagram showing an example of a Golomb-Rice code table.

FIG. 8 is a diagram illustrating an example of an image cycle detection circuit.

FIG. 9 is a block diagram showing an internal configuration example of an image encoding device which is another embodiment of the present invention.

FIG. 10 is a diagram showing a configuration example of a cycle detection unit.

[Fig. 11] Fig. 11 is a diagram for describing an example of a unit configuration and a cutout method for determining an encoding order.

FIG. 12 is a block diagram showing an internal configuration example of an image encoding device according to another embodiment of the present invention.

FIG. 13 is a diagram for explaining another example of the unit cutout method.

FIG. 14 is a block diagram showing an example of the internal configuration of an image encoding device according to another embodiment of the present invention.

FIG. 15 is a diagram showing an example of enlargement processing in the resolution conversion method.

FIG. 16 is a flowchart illustrating an example of an image encoding process to which the present invention is applied.

FIG. 17 is a flowchart for explaining another example of the image coding process to which the present invention is applied.

FIG. 18 is a flowchart for explaining another example of the image encoding process to which the present invention is applied.

FIG. 19 is a flowchart for explaining another example of the image encoding process to which the present invention is applied.

FIG. 20 is a diagram illustrating an example of dither processing using the systematic dither method.

[Explanation of symbols]

10 ... Statistical means, 11 ... Unitizing means, 12 ... Encoding means, 20, 35 ... Cycle detecting means 21, 32, 42 ...
Unitization means, 22, 33, 43 ... Unit cutting means, 23, 34, 44 ... Encoding means, 30, 40 ... Resolution conversion means, 31, 41 ... Gradation processing means, 36 ... Resolution conversion information creating means, 45 ... Judgment means, 46 ... Image cycle information, 47 ... Resolution conversion information.

   ─────────────────────────────────────────────────── ─── Continued front page    F-term (reference) 5C078 AA01 BA22 CA01 CA31 DA01                       DA11 DB16                 5J064 AA02 BA08 BB09 BC23 BD06

Claims (30)

[Claims] 1. Digital image data having periodicity 1
A unit is configured by using a plurality of pixels belonging to a period, the configured unit is used as an encoding unit, and two units that are continuously encoded are separated by an integral multiple of the period. Image encoding device. 2. The image code according to claim 1, wherein the unit to be encoded next by an integral multiple of the period is adaptively determined based on the encoded unit in the digital image data. Device. 3. A pixel belonging to the unit is a black pixel,
2. The image coding apparatus according to claim 1, wherein the occurrence probabilities indicating the likelihood of white pixels being similar to each other. 4. The image coding apparatus according to claim 1, wherein the pixels belonging to the unit have similar pixel value change probabilities between two pixels separated by an integral multiple of the period. . 5. The image coding apparatus according to claim 1, wherein the number of pixels belonging to the unit is not the same in the digital image data. 6. The image coding apparatus according to claim 3, wherein the occurrence probability is obtained by measuring the number of black pixels or white pixels for each phase in a cycle. 7. The change probability is 1 for each phase in a cycle.
The image encoding device according to claim 4, wherein the image encoding device is obtained by measuring the number of pixels of a pixel in which the pixel value of the next pixel separated by a cycle has changed. 8. The image coding apparatus according to claim 5, wherein the number of pixels belonging to the unit is adaptively determined based on a coded unit in the digital image data. 9. The image coding apparatus according to claim 1, wherein the number of pixels belonging to the unit is a divisor of the cycle. 10. The image coding apparatus according to claim 1, wherein coding is performed on the unit. 11. A unit is configured by using a plurality of pixels of digital image data having a periodicity, and the configured unit is used as an encoding unit. In the image coding apparatus separated by N times (N is a natural number), the configuration of the unit is determined based on a cycle of the digital image data, the image coding apparatus. 12. The number of pixels belonging to the unit is M times the cycle of the digital image data, or 1 / M.
The image encoding device according to claim 11, wherein the number is doubled (M is a natural number). 13. The periodic digital image data is obtained by performing gradation processing on digital image data having a predetermined resolution, and the natural number N is determined by the gradation processing. 13. The image coding apparatus according to claim 11, wherein the image coding apparatus is determined by using a subsequent image cycle and resolution information. 14. The natural number N is determined based on the least common multiple of the image period and the number of consecutive equivalence pixels corresponding to the resolution information.
3. The image coding device according to 3. 15. When the least common multiple exceeds a predetermined threshold value Th (Th is a natural number), the number of pixels between two units separated by N times the cycle is set to a value equal to or less than the threshold value Th. The image coding device according to claim 14, wherein 16. A determination unit that determines a predetermined condition based on the image period and the number of consecutive equivalence pixels corresponding to the resolution information, and performs a predetermined conversion based on the result of the determination unit. The image coding apparatus according to any one of claims 13 to 15, characterized in that. 17. The judging means determines, as the predetermined condition, a difference between the number of consecutive equivalence pixels corresponding to the resolution information and P times or 1 / P times (P is a natural number) the image period. The image coding apparatus according to claim 16, wherein it is determined whether or not there is a natural number P that is equal to or less than a predetermined threshold value Ti (Ti is a natural number). 18. The predetermined conversion is performed by converting the number of consecutive equivalence pixels corresponding to the resolution information to P times or 1 / P times (P is a natural number) the image period. Item 17. The image encoding device according to Item 17. 19. A unit is configured by using a plurality of pixels belonging to one cycle of digital image data having periodicity, and the configured unit is used as an encoding unit, and between two units that are continuously encoded. Is an integer multiple of the period, the image coding method. 20. The image code according to claim 19, wherein the unit to be encoded next by an integral multiple of the period is adaptively determined based on the encoded unit in the digital image data. Method. 21. The image coding method according to claim 19, wherein the pixels belonging to the unit have similar occurrence probabilities indicating the likelihood of becoming a black pixel or a white pixel. 22. The image coding method according to claim 19, wherein the pixels belonging to the unit have similar pixel value change probabilities between two pixels separated by an integral multiple of the period. . 23. The image encoding method according to claim 19, wherein the number of pixels belonging to the unit is not the same in the digital image data. 24. The occurrence probability is, for each phase in a cycle,
22. The image encoding method according to claim 21, wherein the image encoding method is obtained by measuring the number of black pixels or white pixels. 25. The change probability is set for each phase in a cycle.
3. It is obtained by measuring the number of pixels of a pixel in which the pixel value of the next pixel that is one cycle apart changes.
2. The image encoding method described in 2. 26. The number of pixels belonging to the unit is adaptively determined based on an encoded unit in the digital image data.
The described image coding method. 27. The image coding method according to claim 19, wherein the number of pixels belonging to the unit is a divisor of the period. 28. The image coding method according to claim 19, wherein the unit is coded. 29. The image coding apparatus according to any one of claims 1 to 18, or a function for realizing a function as each unit of the image coding apparatus, or any one of claims 19 to 28. A program for executing the image coding method described in 1. 30. A computer-readable recording medium in which the program according to claim 29 is recorded.