JPS5826712B2 - Redundancy reduction coding transmission method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention is a method for encoding and transmitting a binary image signal obtained by scanning an information source. The present invention relates to a method of encoding and transmitting pixel information of a second scanning line immediately after a first scanning line based on a portion.

In pixel information having two levels, such as a facsimile signal, there is a strong correlation between pixels in the scanning line direction and a strong correlation between pixels on adjacent scanning lines.

Conventionally, there is an intra-pattern predictive differential encoding method as shown below as a method that also utilizes this type of correlation between scanning lines.

FIG. 1 is a diagram explaining the principle of the intra-pattern predictive differential encoding method.

Hereinafter, the scanning line immediately before the scanning line to be encoded will be referred to as a first scanning line #1, and the scanning line to be encoded will be referred to as a second scanning line #2.

Further, 1.2 are pixels of the second and first levels, respectively.

In this method, as shown in FIG. As shown in FIG. Only one first level run on the scan line and only one on the second scan line
As shown in FIG. A fourth pattern in which two or more first-level runs on a scanning line touch each other, and a fourth pattern in which two or more first-level runs on a second scanning line touch each other, and a single first-level run on a second scanning line and a first scanning pattern as shown in FIG. A fifth pattern in which two or more first level runs on the line touch each other is identified, and each pattern is uniquely encoded and transmitted.

However, on the actual screen, as shown in Figure 2, the above 5
Cases may occur that do not correspond to any of the three patterns.

In this case, in the conventional method, the pattern is cut along the broken line 3, and the patterns are divided into pattern (2) and pattern (2) and then encoded.

This has the disadvantage that the encoding device becomes complicated.

In order to eliminate such drawbacks, the present invention encodes the fifth pattern as the third pattern, inverts the first level of the already encoded pattern to the second level, and then encodes the new pattern. For the fourth pattern, the relative shift from the first scanning line to the second scanning line at the leftmost and rightmost ends of the first level in the pattern is determined by identifying the fourth pattern. 3, and only the state of the branch on the second scanning line is encoded as a fourth pattern.The drawings will be described in detail below.

FIG. 3 is an example of a facsimile image signal, with scanning line #1
is the first scanline already encoded, and scanline #2
is the second constant scan line to be encoded.

4 is the scanning direction, and here it is assumed that scanning is from left to right.

However, this does not result in loss of generality.

A specific example of encoding will be shown below.

Furthermore, the first level is black and the second level is white.

There is no loss of generality in this.

5 and 6 indicate white and black pixels, respectively.

The pixel information of the first scanning line #1 is sequentially read to detect the position pH where the first black level appears.

Furthermore, a point Q11 where this first level ends is detected.

The pixel P12 directly below the edge point P11 is checked, and if the level is black, the level of the pixel on the second scanning line is checked in the left direction to detect the point where the black level ends.

Conversely, if the pixel P1□ is at the white level, the second scanning line #2 is scanned rightward to find the point where the black level appears up to just below the point Q11.

At this time, if the black level does not appear even if it reaches just below point Q,1, then this part is in a state where there is no black run on the second scanning line that is in contact with the black run P11Q11 on the first scanning line ( 2nd pattern).

The example in the figure is this case.

Thereafter, the second scanning line is scanned to detect whether there is a black pixel between the first pixel and pixel P0.2.

In the figure, portions 12 and 14 are black pixel portions, and this state is identified as a state (first pattern) where no black run that is in contact with a black run on the second scan line is on the first scan line. .

The first pattern is encoded as a set of white pixel runs and black pixel runs.

In the figure, there are two first patterns up to pixel P12 on the second scanning line.

One is 4, and 12 silk, and the other is 13 and 44
This is a group of

The respective lengths of 1 and 14 are 3, 2, and 2.1 in order.

Therefore, this part is encoded as follows. Next, identify and encode how the black run of PxxQtt looks on the second scan line.

In the case of the figure, since the PIIQII part is identified as the second pattern as previously investigated, this pattern is encoded as follows.

P(2) However, P(2) is the second pattern identification code.

After this, the pixel Q12 immediately below Qo is detected, which is one pixel to the left of Qo, and then the detection of pixels to the right of Qo is started.

When the black pixel P21 is detected, the ending point Q of the black pixel is further detected.
21 is detected, and it is further detected whether or not there is a run in contact with the black pixel run of P21Q21 in the second scanning line.

In this case, it can be seen that run P22Q22 exists and is not the second pattern.

After detecting the position of pixel P22, Q1□ on the second scanning line
It is detected whether or not the first pattern exists between P22 and P22.

In this case it does not exist.

Next, the black run portion of P21Q21 is encoded as a state (third pattern) in which the black run adjacent to the black run on the first scanning line is on the second scanning line.

Specifically, the leftmost end P21 t P2□ and the rightmost end Q21 t of the respective black levels on the first and second scanning lines.
Detect Q22 and calculate the deviation d1 of the left and right ends from the first scanning line to the second scanning line = P22 - P21td - Q22
Encode Q21.

In the figure, since d=+1°d2=-2, encoding is performed as follows.

Next, it is detected whether a white run exists between P2° and Q2□.

In this case, there is no black pixel, so we start detecting whether there is a black pixel to the right of pixel Q21, and when black pixel P31 is detected,
Next, black pixel end point Q31, then black run IIIQ:
It is detected whether or not there is a black run in contact with 11 on the second scanning line, and if there is a black run in contact with the second scan line, the start point P32 of the first black run and the end point Q3□ of the last run are detected.

Thereafter, it is checked whether there is a first pattern between Q22 and P32, and in this case, since it does not exist, the deviations d3, t, and d4 of P, 111P32, and Qa, tQa2 are each encoded as a third pattern.

In the example shown in the figure, d3=After this, detection is started to determine whether a white run exists between P32 and Q3□, and if a white run exists, a A state in which there are two or more runs on the second scanning line (fourth pattern)
encoded as .

In other words, in this case, the black run P31Q3 on the first scanning line
1 is branched into a plurality of black runs on the second scanning line.

The fourth pattern is encoded as a combination of black runs and white runs.

In the figure, there are two fourth patterns between P32 and Qa2.

One is a pair of 15 and 16, and one is a pair of 17 and 18 silk.

Since 15 to 18 are respectively 2.2, 1, and 3, they are encoded as follows.

P(4)L(2)L(2)P(4)L(1)L<3) However, P(4) is the fourth pattern identification code.

Although this has been omitted so far, if the black pixels in the already encoded pattern are replaced with white pixels, the remaining portion of FIG. 4 becomes as shown in FIG. 5.

After such post-processing, the process moves on to detecting the black pixel on the right side of Q3B.

In this case, since there is no longer a black pixel, a pixel P41 t ” is virtually next to the last pixel on the first and second scanning lines.
4 Think about □.

Next, it is detected whether the first pattern does not exist between Q3° and P42.

In the example shown, '9 t lto is 2°3, so it is encoded as follows.

P ('1) L (2) L (3) In this way, by replacing the black pixels of the already encoded pattern with white pixels and performing the subsequent state identification and encoding, it is possible to perform the following state identification and encoding. This eliminates the need for complex processing such as pattern cutting.

FIG. 6 shows an example of a code P(.) for identifying a pattern.

Even if the third pattern identification code has a statistically high probability of occurrence, it is also shortened.

The relative distance d takes positive, negative, or zero values.

Therefore, cases other than D(0) are further expressed as follows.

An example of codes for D(0) and S(to) is shown in FIG.

Statistically, the deviation value is concentrated around zero, and since it is most often zero, the shortest code is assigned to D(0).

Examples of the code word L(1) representing the length l and the code word DA(ldl) representing the absolute value 1 dl are shown in FIGS. 8 and 9, respectively.
As shown in the figure.

FIG. 10 shows an embodiment of the encoder part of the present invention, in which γ is an input line such as a facsimile signal having two levels, and 8 is a storage of pixel information of the first scanning line for one scanning line. The line memory 1.9 stores the pixel information of the second scanning line as 1.
The line memory 2.10 that stores scanning line data is a line memory address control circuit that controls writing and reading of the line memories 1 and 2.

When the encoding for one scan is completed, the contents of the line memory 2 are transferred to the line memory 1.

Thereafter, pixel information for one new scanning line from the facsimile terminal is stored in the line memory 2 via the signal line T.

Before explaining the operation in FIG. FIG. 11 shows reference pixel positions in the third and fourth patterns.

FIG. 11a shows an example of the third and fourth patterns, where Pl is the starting point Q1 of the black run on the first scanning line and the ending point, and Pl is the starting point Q1 of the black run on the first scanning line, and Pl is the line memory 2 in contact with the black run P1Q1. The starting point of the first black run, Ql is the end point of the last black run on line memory 2 that is in contact with black run PIQ1, and Q3 is the end point of the last black run on line memory 1 that is in contact with the black run that includes point Q2. It is the end point.

Note that Ql and Q3 may match.

In the case of the second pattern, the previous pixel immediately below Ql is defined as Ql.

Furthermore, Ql and Q3 match.

The explanation of the operation shown in FIG. 10 will be continued below.

The line memory address control circuit 10 reads the pixels on the line memory 1 from immediately after Q3 of the pattern encoded immediately before, and the Pl, Q1 detection circuit 11 reads the pixels of the new pattern P.
l Furthermore, Ql is detected.

After that, read out the pixels on line memory 2 and P
l.

Detect Ql.

If Pl is not detected, a second pattern has been identified and this is stored in the second pattern encoder 14.

Pl. The Q2 detection circuit has the point Q of Ql of the previous pattern.
2' is stored, the line memory address scans Q2' to P2 on the second scanning line, and the read pixel information is sent to the first pattern encoder 13.

The first pattern encoder 13 includes a run-length encoder, encodes the first pattern every time it detects a black run, and sends the codes to the signal line 1γ in a determined order.

When the first pattern does not exist or when the encoding of the first pattern is completed, the second encoder 14 checks whether the second pattern has been detected, and if detected, encodes the second pattern. , the code is sent to the signal line 1γ.

If the second pattern is not detected, the third pattern is encoded by third pattern encoding 15.

The third pattern encoder 15 calculates the differences between Pl and Pl, and between Ql and Ql, encodes each difference, and sends them to the signal line 11.

Next, the line memory address control circuit 10 reads pixels between 12 and 02 on the line memory 2 and sends them to the fourth pattern encoder 16. Perform encoding of the fourth pattern,
The codes are sent to the signal line 1γ in the determined order.

If there is no fourth pattern, or if the encoding of the fourth pattern is finished, the Pl, Q1 detection circuit will detect Q with Ql as a reference.
Detection 3 is performed.

When the detection of Q3 is completed, the line memory address control circuit 10 sets the pixel information after Q3 in the line memory 1 to Pl.
.

The signal is sent to the Q1 detection circuit 11 to detect a new Pl, and the same process is repeated until all encoding in the line memory 2 is completed.

When the encoding of the line memory 2 is completed, the contents are transferred to the line memory 1, pixel information of one new scanning line is stored in the line memory 2, and the encoding of the line memory 2 is repeated again.

Note that when encoding the first pattern, as long as the signals between Q2' and P2 of the line memory 2 are read and processed, the operation of replacing the black pixels of the already encoded pattern with white pixels is uniformly distributed. It becomes unnecessary for the song.

Next, a method for limiting the deterioration of the image quality of the reproduced screen due to code errors, which is another object of the present invention, will be described.

In the encoding method of the present invention, in order to encode the image signal of the second scanning line, image signal information of the first scanning line immediately before the second scanning line is used to encode the second scanning line. is being carried out.

Therefore, the decoder side also decodes the image signal of the second scanning line using the image signal information of the first scanning line that has already been decoded.

In this way, since encoding and decoding are performed while sequentially using the image signal information of the immediately preceding scanning line, coding errors may occur due to the influence of line noise, and the image signal of a certain scanning line may not be reproduced correctly. Otherwise, all the image signals of the scanning lines subsequent to this scanning line will not be reproduced correctly, and the image quality of the reproduced screen will deteriorate significantly.

Therefore, in order to detect the occurrence of a code error, to suppress the deterioration of the image quality of the scanning line where the code error has occurred, and to prevent the deterioration of image quality due to the code error from spreading beyond the scanning line, it is necessary to immediately correct the code error state. Restoration will be required.

According to the present invention, for these purposes, the encoder side inserts a so-called self-synchronized second control code that can be detected from any location in the code sequence into the image signal in a predetermined interval, for example, K( K=1,2.

3,...) Periodically inserted immediately before the start of encoding of the first /161 scanning line of each scanning line, and for the /161 scanning line, the image signal information of the immediately preceding scanning line is not used. The image signal information of the /I61 scanning line is encoded, and the /161
The scanning lines 42, /163...4 immediately after the scanning line are encoded according to the present invention, and a first control code for detecting the occurrence of a code error is inserted every time the encoding of each scanning line is completed. .

Note that the first control code is also inserted immediately after the encoding of the /16.1 scanning line is completed.

On the decoder side, when the self-synchronized first control code is detected, the immediately following code sequence is assumed to be PL encoded and the information of the immediately preceding scanning line is not used (A
Decoding is performed as one scanning line.

Thereafter, it is assumed that encoding according to the present invention has been performed, and decoding is performed using the immediately preceding scanning line information.

Every time the decoding of one scanning line is completed, the first control code is decoded for the immediately following code, and the presence or absence of an error is checked.

If an error is detected, the decoded scan line and subsequent scan lines are replaced with the image signal of the scan line immediately before the decoded scan line to suppress deterioration in image quality.

Moreover, when the self-synchronized second control code is detected, decoding of the RL is immediately started to recover from the error state.

FIGS. 12 and 13 show an encoder according to a second embodiment of the present invention,
FIG. 2 is a block diagram of a decoder.

The facsimile image signal input line γ is connected to the RL encoder every K scanning lines via a switch 100 under the control of a switch control circuit 101.

At this time, the second control code generation circuit 104 generates a second control code, and then the RL encoder 102 RL encodes the scanning line (waste 1 scanning line), and the switch 100 is connected to the encoder 103 according to the present invention. , A2 to AK scanning lines are encoded according to the present invention, and a first control code is inserted by the first control code insertion circuit 105 every time the encoding of each scanning line is completed.

On the decoder side in FIG. 13, the second control code detection circuit 106
When the second control code is detected, the RL decoder 10r decodes the RL for one scanning line (one scanning line).
Write the reproduced pixel information to the line memory 3 of 108, /
V) When the decoding of one scanning line is completed, the contents of the line memory 3 are transferred to the line memory 4 of 109.

Thereafter, using the contents of the line memory 4, the decoder 110 according to the present invention sequentially processes A62. /%3...Decodes the AK scanning line.

Note that each time the decoding of each scanning line is completed, the first control code is detected by the first control code detection circuit 10γ, and the code error detection circuit 112 detects whether or not a code error has occurred.

Once a code error occurs, the contents of the line memory 4 are sent out without decoding the AK scanning lines after that scanning line.

After that, when the second control code is detected, normal decoding operation is started and the code error state is recovered.

As described above, according to the present invention, a state in which a black run on a first scanning line branches into two or more black runs on a second scanning line is identified, and The leftmost and rightmost edges of each black level on the scanning line are detected, the relative shift between the left and right edges from the first scanning line to the second scanning line is encoded as a first function, and the relative shift on the second scanning line is encoded as a first function. The branch state of the run is encoded as a second function, the black pixels in the already encoded part are replaced with white pixels, and the new state is identified. This makes effective use of the correlation between scanning lines, resulting in high efficiency. This method has the advantage that it is possible to perform redundancy suppression coding and that the coding can be performed using a simple circuit.

For example, after inserting a self-synchronized second control code for each scanning line, RL encoding is performed for only one scanning line, and the encoding of the present invention is performed for the subsequent scanning lines. By inserting the first control code for code error detection each time the coding is completed, there is an advantage that the spread of image quality deterioration due to code errors can be prevented and recovery from a code error state can be made promptly.

Furthermore, it goes without saying that "transmission" in this specification includes the transfer of information within the same device.

[Brief explanation of drawings]

FIG. 1 is an explanatory diagram of identified patterns in the conventional encoding method, FIG. 2 is an explanatory diagram of pattern cutting in the conventional encoding method, and FIG. 3 is an explanatory diagram for explaining the principle of encoding of the present invention. FIG. 4 is an explanatory diagram showing an example of encoding according to the present invention, and an explanatory diagram showing the encoded signal of FIG. 3. FIG. 5 is a supplementary diagram for explaining the principle of encoding according to the present invention. , 6th to 9th
The figure is an explanatory diagram showing an example of code assignment for each encoding function, Figure 10 is a block diagram showing an example of the encoder according to the present invention, and Figure 11 defines points that serve as the reference for each pattern in the present invention. 12 is a block diagram showing another embodiment of the encoder of the present invention, and FIG. 13 is a block diagram showing an embodiment of the decoder of the present invention. 1... Second level pixel, 2... Second level pixel, 3... Cutting broken line, 4... Scanning direction, 5... White pixel, 6... Black pixel.

Claims (1)

[Claims] One pixel is scanned and decomposed into pixels of a first level and a second level, and the first scanning line immediately before the scanning line to be encoded and the second scanning line to be encoded are separated. In a redundancy reduction coding transmission method in which the state of a pixel area having a first level is encoded and transmitted between no first level runs on the first scan line that touch a first level run on the second scan line; and no single first level run on the first scan line. A means for identifying a state in which a first level run on a second scanning line is in contact with a first level run, and a means for encoding and transmitting each state identified by the means are provided. and means for identifying a condition in which one first level run on a first scan line branches into two or more first level runs on a second scan line; 1. 2nd
means for detecting the leftmost and rightmost ends of each first level on the scan line and encoding the relative shift of the left and right ends from the first scan line to the second scan line as a first function; , and means for encoding the state of a branch of a run on a second scan line as a second function, and a new state after replacing the first level of the already encoded and transmitted state with a second level. A redundancy suppression coding transmission system characterized by providing means for identifying and coding. 2 The screen is scanned and decomposed into pixels of the first level and the second level, and between the first scanning line immediately before the scanning line to be encoded and the second scanning line to be encoded, the first In a redundancy suppression coding transmission method that encodes and transmits the state of a pixel area having a level of no first level runs on the first scan line that touch the first level runs on the second scan line, and a single first level run on the first scan line. means for identifying the states in which the runs of the first level on the second scanning line are in contact with each other, and means for encoding and transmitting each state identified by the means; means for identifying a condition in which one first level run on a line branches into two or more first level runs on a second scan line;
means for detecting the leftmost and rightmost ends of each first level on the scan line and encoding the relative shift of the left and right ends from the first scan line to the second scan line as a first function; , and means for encoding the state of a branch of a run on a second scan line as a second function, and a new state after replacing the first level of the already encoded and transmitted state with a second level. Furthermore, an image signal having two levels obtained by scanning the screen is divided into predetermined sections, and an error occurs synchronously immediately after the end of encoding of the section. Immediately before encoding the first control code for detection and every K (K=1, 2, 3...) section, insert a self-synchronizing second control code for recovery from an error state. and for the first section of each predetermined section, encoding the section using only the image signal information of the section without using the image signal information of the sections before the section. 1. A redundancy suppression coding transmission system characterized by providing means for performing redundancy suppression coding.