JPH05145767A - Data compressing method - Google Patents

Abstract

PURPOSE:To decrease the number of bits and to improve the compression efficiency by executing modified Huffman MH encoding, when one of image data of the previous and the present lines is full white, and the other is non-full white. CONSTITUTION:A control part 5 decides whether a modified read MR encoding continuous limit number K and an MR encoding continuous number L coincide with each other or not, and when both of them coincide with each other, picture data of an object line is subjected to MH encoding by an MH encoding part 6, subjected to code output by adding an end-of-line EOL and a one-dimensional tag of a code generating part 5 and the continuous number L is reset to '0', and the next line data is fetched. Also, when it is decided that one of the previous line and the object line is full. white and the other is non-full white, since there is no correlation, the control part 3 executes MH encoding in the same way. Subsequently, when it is decided that both of them are full white, an MR encoding line is continued. On the other hand, when both of them are non-fun white, encoding is executed by the object line data encoding part 6, the EOL and a two-dimensional tag of the generating part 5 are added and outputted, the continuous number L is updated to L+1, and the next line data is fetched and processed in the same way.

Description

Detailed Description of the Invention

(Table of Contents) Industrial Application Field of the Prior Art (FIGS. 6 to 7) Problem to be Solved by the Invention (FIG. 8) Means for Solving the Problem (FIG. 1) Action Embodiment (a) Description of Embodiment 1 (FIGS. 2 to 4) (b) Description of Second Embodiment (FIG. 5) (c) Description of Other Embodiments Effect of the Invention

[0002]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention sets the image data of the current line to a pass code, a horizontal code, a horizontal code, in accordance with the relative position between the changing point of the image data of the previous line and the changing point of the image data of the current line.
The present invention relates to a data compression method for performing MR coding using a vertical code.

Image data has an enormous amount of data as compared with code data, and various compression techniques utilizing the characteristics of images have been adopted. Data is compressed by a facsimile device, an image file device or the like. It communicates and remembers.

Such data compression techniques include a one-dimensional encoding method and a two-dimensional encoding method. In a facsimile apparatus, G3 CCITT Recommendation T.264. In 4, the former is MH (Modified Hoffman) code and the latter is MR
The (Modified Read) code is recommended.

This MR coding has high compression efficiency and is widely used, but further improvement in compression rate is required.

[0006]

2. Description of the Related Art FIGS. 6 and 7 are explanatory views (1) and (2) of a conventional technique. As shown in FIG. 6 (A), M
In the R coding method, the current line is coded by using the previous line as a reference line, the current line as a coding line, a ₀ as a reference or starting point change pixel on the coding line, and a ₁ as a code. On the encoding line, the first change pixel on the right of a _0, on the encoding line, a ₂ is the first change pixel on the right of a ₁ , and b ₁ is the change pixel on the reference line.
As a first pixel, the right change pixels b ₁ to b ₂ on the reference line of opposite colors with a ₀ in the right of a _0, to encode.

As shown in FIG. 7 (A), the coding modes include a pass mode, a vertical mode, and a horizontal mode. The pass mode is, as shown in FIG. 6 (B), on the left side of a ₁ , b
It is defined by the presence of ₂ and when this mode is coded, a ₀ is set on the pixel (a ₀ ′) on the coding line immediately below b ₂ in preparation for the next coding.

In the vertical mode, as shown in FIG.
The position of a ₁ is encoded by the relative position from b ₁ , and the relative distance is represented by seven different values, V0 and V.
It takes any value of R1, VR2, VR3, VL1, VL2, and VL3, and in the example in the figure, is encoded by VL2, and after vertical mode encoding, the position of a ₀ is moved to a ₁ .

In the horizontal mode, as shown in FIG.
Both run lengths a ₀ a ₁ and a ₁ a ₂ are coded H +
Encoding is performed using M (a ₀ a ₁ ) + M (a ₁ a ₂ ).
H is a flag code in the two-dimensional code table (FIG. 7 (A)), and M (a ₀ a ₁ ) and M (a ₁ a ₂ ) are each run a _0.
It is a code indicating the length and color of a ₁ and a ₁ a ₂ , and the position of a ₀ is moved onto a ₂ after horizontal mode coding.

The MR encoding procedure is performed by first
Mode, then the relative distance a ₁b₁Find the absolute value of
Therefore, if this absolute value is less than or equal to “3”, then according to FIG.
A ₁b₁Is encoded in vertical mode and exceeds "3"
For example, according to FIG. 7 (A), the horizontal mode code is ‘001’.
Continuing, a₀a₁And a₁a₂To each one-dimensional encoding
Encode more.

In such an MR encoding method, since the reference line is used as a reference, even a 1-bit error has a large influence on the subsequent line. Therefore, as shown in FIG.
As shown in (C), an MH coding line which is one-dimensional coding is inserted into every four MR coding lines, and M
The influence of errors on the R coding line is minimized.

[0012]

FIG. 8 is a diagram for explaining the problems of the prior art. By the way, this MR encoding method is most efficient when the image data of the previous line and the image data of the current line are the same, and is efficient when there is no correlation between the image data of the previous line and the image data of the current line. Is bad.

The correlation between the image of the previous line and the image of the current line is difficult to judge at first glance, but a general manuscript has a line where characters are present and a blank (all white) upper / lower part. It consists of margins and line spacing.

When this is examined, as shown in FIGS. 8A and 8B, when the previous line is all white and the current line is not all white, or when the previous line is not all white and the current line is all white. Is all white, the MR encoding is in horizontal mode or pass mode,
The MH coding method, which is a one-dimensional coding method, requires a smaller number of bits.

Further, as shown in FIGS. 8C and 8D, when the previous line and the current line have the same pattern, the vertical code V0 is sufficient in the MR coding method, and especially the previous line and the current line. Is all white, one vertical code V0 (1 bit)
The number of bits is much smaller than that of the MH coding method.

For this reason, in the prior art, MH coding is performed for every constant number of lines regardless of the correlation between the image of the previous line and the image of the current line, so that there is a problem that the compression efficiency cannot be effectively improved. It was

Therefore, the present invention provides a data compression method capable of improving compression efficiency by selecting MR coding or MH coding depending on the presence or absence of correlation between the image of the previous line and the image of the current line. The purpose is to

[0018]

FIG. 1 shows the principle of the present invention. According to claim 1 of the present invention, the image data of the current line is converted into a pass code, a horizontal code, and a vertical code according to the relative position of the change point of the image data of the previous line and the change point of the image data of the current line. In a data compression method for MR coding using
It is detected that one of the image data of the previous line and the image data of the current line is all white and the other image data is non-white, and the image data of the current line is changed to black. Alternatively, it is characterized in that MH coding is performed with a code indicating a white run length.

A second aspect of the present invention is based on the first aspect.
When the MR encoded line continues for a predetermined number of lines, the image data of the next line is MH encoded.

According to a third aspect of the present invention, the image data of the current line is converted into a pass code, a horizontal code, and a horizontal code according to the relative position between the change point of the image data of the previous line and the change point of the image data of the current line. A data compression method for performing MR coding using a vertical code and subjecting the image data of the next line to MH coding with a code indicating a black or white run length by the MR coding line continuing for a predetermined number of lines. The image data of the previous line and the image data of the current line are detected to have the same pattern, and the image data of the current line is MR-coded.

According to a fourth aspect of the present invention, in the third aspect,
Detecting that one of the image data of the previous line and the image data of the current line is all white, and the other image data is non-all white, the image data of the current line is
MH encoding is performed with a code indicating a black or white run length.

According to a fifth aspect of the present invention, in the third or fourth aspect, it is detected that the image data of the previous line and the image data of the current line have the same pattern, and the image data of the current line is detected. Since the MR coding is performed, the continuous allowable number of the MR coded lines having the same pattern is increased.

A sixth aspect of the present invention is characterized in that, in the third, fourth or fifth aspect, the same pattern is an all white pattern.

[0024]

According to the first aspect of the present invention, as shown in FIGS. 8A and 8B, the image data of either the image data of the previous line or the image data of the current line is entirely white. , When the other image data is non-white, there is no data correlation and MR
Since the number of bits is smaller in MH encoding than in encoding, MH encoding is performed instead of MR encoding, and MH encoding also has an effect of not impairing recoverability against errors.

According to a second aspect of the present invention, the MH according to the first aspect.
If the number of coding lines is small, the recoverability is deteriorated. Therefore, the MR coding line continues for a predetermined number of lines, so that the image data of the next line is forcibly MH coded,
We improved the recoverability.

In claim 3 of the present invention, as shown in FIG.
As shown in (D), when the image data of the previous line and the image data of the current line have the same pattern, the data is correlated, so that the MR coding is more suitable than the MH coding. Since the number is small, change to MH coding,
Even if an error occurs due to the MR coding, even if an error occurs due to the same pattern, an output equivalent to that normally restored can be obtained by replacing the error line with the previously restored line. ..

According to claim 4 of the present invention, one of the image data of the preceding line and the image data of the current line of claim 1 is completely white, and the other image data is non-white. And MH-encoding the image data of the current line with a code indicating a black or white run length,
The compression efficiency was improved.

According to the fifth aspect of the present invention, when the image data of the previous line and the image data of the current line have the same pattern, even if an error occurs, the error line is replaced with the previously restored line. If no error appears in the output,
Since no apparent error occurs even if the MH coded line is not inserted in principle, the MR having the same pattern is used.
The number of continuous encoding lines is increased to improve the compression efficiency.

According to a sixth aspect of the present invention, if the same pattern is an all-white pattern, only the lines of all-white portions that are likely to occur relatively stably (continuously) are continuously formed. , MR coding is possible.

[0030]

Embodiments (a) Description of First Embodiment FIG. 2 is a configuration diagram of an embodiment of the present invention, showing a configuration of a compression unit.

In the figure, reference numeral 1 denotes an all-white detection section, which is composed of a change point detection circuit and which has image data for one line (for example, 17
28-bit) change point is detected, and when the first bit is white (“0”) and there is no change point, an all-white line notification is generated. Reference numeral 2 is a line buffer. Which temporarily stores the image data of 3, the control unit 3, which performs the encoding process shown in FIG. 3, the storage unit 4 which stores whether the previous line is completely white or not. Is.

Numeral 5 is a code generator, which generates an end-of-line EOL and a one-dimensional or two-dimensional tag under the control of the controller 3, and numeral 6 is an MH encoder. By control, one line of image data in the line buffer 2 is MH encoded, and 7 is an MR encoding unit.
Under the control of the control unit 3, the image data of one line in the line buffer 2 is MR encoded.

The code output is M for each line.
An H code and a one-dimensional tag, EOL or MR encoding and a two-dimensional tag, and EOL are output. FIG. 3 is a processing flowchart of the first embodiment of the present invention, and FIG. 4 is an explanatory view of the first embodiment of the present invention.

The control unit 3 controls the MR coding continuous limit number K.
And the number L of continuous MR encoding are set to "4". This limits the number of consecutive MR encodings to 4.

The control unit 3 controls the MR coding continuous limit number K.
And the number of consecutive MR encodings L match,
If the MR-encoded consecutive number K and the MR-encoded consecutive number L match, an MH encoded line is inserted, and the process proceeds to step.

If the MR-encoded continuous number K and the MR-encoded continuous number L do not match, the control unit 3 outputs the output of the all-white detection unit 1 of the target line, although it is an MR encoded line. By comparing the contents of the previous line of the storage unit 4 whether it is all white or not, it is checked whether either the previous line or the target line is all white and the other is not.

When the control section 3 determines that either the previous line or the target line is completely white and the other is not, there is no correlation as shown in FIGS. 8 (A) and 8 (B). Therefore, in order to perform the MH coding instead of the MR coding, the process proceeds to the step, and if it is not determined that either the previous line or the target line is all white and the other is not, the MR
Proceed to step for encoding.

Next, the control unit 3 determines whether both the previous line and the target line are all white, and when it determines that both the previous line and the target line are all white, it determines the MR encoded line to be more white. In order to make it continuous, the process proceeds to step, and if it is not determined that both the previous line and the target line are all white, the image data of the target line is MR-encoded by the MR encoding unit 6, and the EOL of the code generation unit 5 and A dimension tag is added, code output is performed, the MR encoded continuous number L is updated to L + 1, image data of the next line is extracted, and the process returns to step.

The image data of the target line is MH-encoded by the MH encoder 6, the EOL of the code generator 5 and a one-dimensional tag are added, the code is output, and the MR encoding continuous number L is "0".
, The image data of the next line is taken out, and the procedure returns to step.

The control unit 3 sets the MR encoded continuous number K to "X"(X> 4) and the MR encoded continuous number L to "0". As a result, the MR-encoded consecutive number K increases, and the MR-encoded consecutive number L is initialized regardless of the previous consecutive MR-encoded number.

The control unit 3 determines from the output of the all-white detection unit 1 whether the target line is all-white. If the line is not all-white, the process returns to step. If it is all-white, the process proceeds to step. The control unit 3 determines whether or not the MR-coded continuous limit number K and the MR-coded continuous number L match, and the MR-coded continuous limit number K and the MR-coded continuous number L match. Is
In order to insert the MH encoded line, the image data of the target line is MH encoded by the MH encoding unit 6, the EOL of the code generation unit 5 and the one-dimensional tag are added, the code is output, and the image of the next line is output. Retrieve the data and return to step.

On the contrary, the MR encoding continuous limiting number K and the MR
If the number of consecutive encodings L does not match, the image data of the target line is MR-encoded by the MR encoding unit 6, the EOL of the code generation unit 5 and the two-dimensional tag are added, and the code is output.
The encoding continuous number L is updated to L + 1, the image data of the next line is taken out, and the process returns to the step.

In this way, when either the previous line or the target line is completely white with respect to the line to be MR-encoded, and the other is not, as shown in FIGS. 8A and 8B. In addition, since the number of bits decreases with MH encoding,
Instead of MH coding.

If it is determined that both the previous line and the target line are all white, the number of bits is smaller in MR coding as shown in FIG. 8D, so that the previous line and the target line are reduced. As long as the conditions of all white and white continue, MR coding is performed, and the limiting number is controlled by the set value “X” of the MR coding continuous limiting number K, and unlimited is also possible.

In the example shown in FIG. 4, the conventional MR coded line inserts one MH coded line for every four lines, while the previous MR coded lines such as lines 4, 8 and 9 are inserted. Alternatively, if only one of the current lines is all white, it is MH coded instead of MR coded.
If the previous line and the current line are all white, MR encoding will continue, and the compression efficiency will be greatly improved.

Also in this case, if the MH coded line is properly inserted, and if both the previous line and the current line are all white, error recovery is easy, so even if MR coding is continued,
The error recovery performance does not deteriorate.

(B) Description of the Second Embodiment FIG. 5 is a flow chart of the processing of the second embodiment of the present invention, in which the white condition is the same for both the previous and current lines, and the same pattern is used for the previous and current lines. It is a certain condition.

Therefore, in addition to the all-white detecting section 1 in FIG. 2, a coincidence detecting section for the previous line and the current line may be provided. The control unit 3 sets the MR encoded continuous number K and the MR encoded continuous number L to "4".

As a result, the number of consecutive MR encodings is limited to 4. The control unit 3 determines whether or not the MR-coded continuous limit number K and the MR-coded continuous number L match, and the MR-coded continuous limit number K and the MR-coded continuous number L match. Is
To insert the MH coded line, proceed to step.

When the MR-encoded consecutive number K and the MR-encoded consecutive number L do not match, it is an MR-encoded line, but the control unit 3 outputs the same as the output of the all-white detecting unit 1 of the target line. By comparing the contents of the previous line of the storage unit 4 whether it is all white or not, it is checked whether either the previous line or the target line is all white and the other is not.

When the control unit 3 determines that one of the previous line and the target line is completely white and the other is not, there is no correlation as shown in FIGS. 8 (A) and 8 (B). Therefore, in order to perform the MH coding instead of the MR coding, the process proceeds to the step, and if it is not determined that either the previous line or the target line is all white and the other is not, the MR
Proceed to step for encoding.

Next, the control unit 3 determines whether the patterns of the previous line and the target line match, and when it determines that the patterns of the previous line and the target line match, the MR coding is performed. In order to make the lines more continuous, the process proceeds to step, and when it is determined that the previous line and the target line do not match, the image data of the target line is set to the MR encoding unit 6
MR coding is performed, the EOL of the code generation unit 5 and a two-dimensional tag are added, the code is output, the MR coded continuous number L is updated to L + 1, the image data of the next line is extracted, and the process returns to step.

The image data of the target line is MH-encoded by the MH encoder 6, the EOL of the code generator 5 and a one-dimensional tag are added, the code is output, and the MR encoding continuous number L is "0".
, The image data of the next line is taken out, and the procedure returns to step.

The control unit 3 sets the MR encoded continuous number K to "X"(X> 4) and the MR encoded continuous number L to "0". As a result, the MR-encoded consecutive number K increases, and the MR-encoded consecutive number L is initialized regardless of the previous consecutive MR-encoded number.

The control unit 3 determines whether the patterns of the previous line and the target line match, and when it determines that they do not match, returns to the step. When it determines that they match, it advances to the step.

The control unit 3 controls the MR coding continuous limit number K.
And the number of consecutive MR encodings L match,
When the MR-encoded consecutive number K and the MR-encoded consecutive number L match, the MH encoded line is inserted, so that the image data of the target line is MH encoded by the MH encoding unit 6.
The EOL of the code generation unit 5 and the one-dimensional tag are added, the code is output, the image data of the next line is extracted, and the process returns to the step.

On the contrary, the MR coding continuous limiting number K and the MR
If the number of consecutive encodings L does not match, the image data of the target line is MR-encoded by the MR encoding unit 6, the EOL of the code generation unit 5 and the two-dimensional tag are added, and the code is output.
The encoding continuous number L is updated to L + 1, the image data of the next line is taken out, and the process returns to the step.

Even in this case, the same effect as the first embodiment can be obtained. (c) Description of Other Embodiments In addition to the above embodiments, the present invention can be modified as follows.

In the above-described embodiment, the detection unit detects that both the previous and current lines are all white, and that both the previous and current lines have the same pattern.
The control unit 3 detects that only one vertical code V0 or two or more vertical codes V0 are coded, and the coded output is all white on both the previous and current lines, or It may be possible to detect that both of the current lines have the same pattern, in which case it can be detected simultaneously with encoding,
The coding time can be shortened.

The compression unit shown in FIG. 2 is shown as a block.
This may be executed by a microprocessor program. If necessary, either the previous or the current line is entirely white only, the encoding selection control is that both the previous and current lines are all white, both the previous and current lines are the same. It is possible to apply only the encoding selection control depending on the pattern of, and to combine these.

The present invention has been described above with reference to the embodiments.
Various modifications are possible within the scope of the invention, and these modifications are not excluded from the scope of the invention.

[0062]

As described above, according to the present invention,
It has the following effects. In data compression by the MR encoding method, MR encoding and M encoding are performed according to the correlation between the image of the previous line and the image of the current line.
Since the H coding is selected, it is possible to improve the compression efficiency using the correlation of images.

Even in this case, since the MH coded line is properly inserted or the same pattern is used, the recovery performance against errors does not deteriorate.

[Brief description of drawings]

FIG. 1 is a principle diagram of the present invention.

FIG. 2 is a configuration diagram of an embodiment of the present invention.

FIG. 3 is a processing flowchart of the first embodiment of the present invention.

FIG. 4 is an explanatory diagram of the first embodiment of the present invention.

FIG. 5 is a processing flowchart of the second embodiment of the present invention.

FIG. 6 is an explanatory diagram (1) of a conventional technique.

FIG. 7 is an explanatory diagram (part 2) of the conventional technique.

FIG. 8 is a diagram illustrating a problem of the conventional technique.

[Explanation of symbols]

 1 Total White Detection 2 Line Buffer 3 Controller 4 Storage 5 Code Generator 6 MH Coding 7 MR Coding

Claims (6)

[Claims] 1. The image data of the current line is MR-processed by using a pass code, a horizontal code, and a vertical code according to the relative position of the change point of the image data of the previous line and the change point of the image data of the current line. In the data compression method for encoding, it is detected that one of the image data of the previous line and the image data of the current line is all white, and the other image data is not all white, A data compression method characterized in that the image data of the current line is MH-encoded with a code indicating a black or white run length. 2. The data compression method according to claim 1, wherein the image data of the next line is MH-encoded when the MR encoded line continues for a predetermined number of times. 3. The image data of the current line is MR-processed by using a pass code, a horizontal code, and a vertical code according to the relative position of the change point of the image data of the previous line and the change point of the image data of the current line. In the data compression method, the image data of the previous line is MH-encoded by encoding and the image data of the next line is MH-encoded with a code indicating a black or white run length when the MR encoded line continues for a predetermined number of lines. And the image data of the current line are detected to have the same pattern, and the image data of the current line is MR-encoded. 4. The current line is detected by detecting that one of the image data of the previous line and the image data of the current line is all white, and the other image data is non-white. Image data of M with a code indicating the black or white run length.
4. The data compression method according to claim 3, wherein H coding is performed. 5. The image data of the previous line and the image data of the current line are detected to have the same pattern, and the image data of the current line is MR-coded so that the image data of the same pattern is obtained. The data compression method according to claim 3 or 4, characterized in that the maximum number of consecutive MR encoded lines is increased. 6. The data compression method according to claim 3, wherein the same pattern is an all-white pattern.