JPS5813069B2 - Binary signal encoding transmission method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for encoding and transmitting a binary image signal obtained by scanning an information source. The present invention relates to a binary signal encoding transmission method in which pixel information of second and third scanning lines is encoded and transmitted immediately after the first scanning line.

Image signals in facsimiles, etc. contain a lot of redundancy, but in the case of binary signals, the run length of the white or black level is divided into 2 as a method to suppress such redundancy, encode it, and transmit it. A so-called run-length encoding method in which data is encoded and transmitted is conventionally known.

However, such run-length encoding methods only utilize correlation in the scanning line direction and do not utilize correlation between scanning lines, so redundancy cannot be suppressed sufficiently and effective encoding cannot be achieved. There were drawbacks.

In order to eliminate the drawbacks of the conventional example, the present invention provides a level change position on the first scanning line that has already been encoded, and a first
The second and third scanning lines are combined at once by using one of the mode change positions immediately before the mode change position on the second and third scanning lines immediately after the scanning line, which is a group of one pixel each, as the reference coordinate. The present invention provides a binary signal encoding transmission system that performs encoding.

Hereinafter, embodiments will be described in detail with reference to the drawings.

FIG. 1 shows the pixel array to be encoded in order to explain the principle of the present invention. A1 is the first scan line that has already been encoded, and A2 and A3 are the pixel arrays that are to be encoded immediately after the first scan line A1. Second and third scanning lines to be scanned, B indicates a black level pixel, W indicates a white level pixel, and C indicates the main scanning direction.

In addition, one pixel each on the upper and lower sides on the second and third scanning lines A2 and A3,
For example, Q2 and Q4 represent the modes at coordinate P1.

For Q2 and Q4, the white and white mode is mode 0, the white and black mode is mode 1, the black and white mode is mode 2, and the black and black mode is mode 3.

Further, the coordinate position P1 is the position where the mode has changed in the main scanning direction, that is, the position where the mode has changed from mode 0 to mode 1, and is the position to be encoded from now on.

P2 is a position where the mode changes on the second and third scanning lines A2 and A3, and is a position where the mode changes immediately before P1.

P3 is a position at which the pixel level changes on the first scanning line A1, and is the first position at which the pixel level changes after P2 in the main scanning direction.

P4 is the position at which the pixel level changes on the first scanning line, and is the first position at which the pixel level changes after P3 in the scanning direction.

Next, the direction of level change at P1 is determined as follows.

First, the pixel level does not change from white to white from Q1 to Q2 on the second scanning line A2, and Q3 on the third scanning line A3.
From Q4, there is a change in pixel level from white to black.

Therefore, the level change in the scanning line direction at P1 is assumed to be from white to black.

Further, the level change at P3 on the first scanning line A1 is from black to white, and the level change at P4 is from white to black.

Therefore, the position having the same level change direction as P1 is P4
, and this is called P5.

Further, the relative coordinates 2 between P1 and P2 are 1=P1-P2, which is Z25 in FIG.

Note that l is always an integer greater than or equal to 1.

Also, the relative coordinates d between P1 and P5 are P1-P5, and the first
In the figure, d=-3.

Here, the smaller absolute value of ι and d is d, and therefore P5 is selected as the reference coordinate.

Note that d takes a positive, zero, or negative value.

After determining the reference coordinates, the relative coordinates between the reference coordinates and P1 are encoded.

In this example, d=-3 is encoded, and this code is D(
−3).

After that, the mode in P1 is encoded.

In addition, there are two ways to encode P1: one is to directly encode the type of mode in P1, and the other is to directly encode the mode type in P1, and as shown in FIG.
A method of encoding the mode change to 2 is also possible, but
In this example, mode changes, or mode transitions, are encoded.

This mode transition code is expressed as T (m1/m2), which indicates that there has been a transition from mode m1 to mode m2.

Further, in this example, m1=0, m2=1, and the mode transition code is T(0/1).

Therefore, the mode change position P1 is D(-3)T(0/1)
is encoded as

In addition, relative coordinate 2 when P2 is selected as the reference coordinate
The code word for is expressed as L(ι).

When the encoding of P1 is completed in this way, the mode change position P1' immediately after P1 is encoded next.

At this time, the reference coordinate candidates P2, P3, P4, P
5 are respectively P1, P4, P4', and P4 in FIG.
Therefore, the reference coordinate becomes P1, and therefore the code word of P1' becomes L(1)T(1/3).

Such processing is sequentially repeated until the encoding of the positions of all mode change points on the second and third scanning lines is completed.

Once the encoding of the second and third scanning lines is completed, the third scanning line A3 is regarded as the new first scanning line, and the fourth and fifth scanning lines are regarded as the above-mentioned second and third scanning lines. Similarly, encode
Such processing is performed for all scanning lines of one screen.

Note that when encoding the first scanning line and the second scanning line of the screen, there is no particular problem even if the first scanning line A1 serving as a reference is all regarded as white pixels.

Next, a specific example of code assignment will be explained.

In the code word D(d), d takes a positive, zero, or negative value,
Moreover, its distribution peaks at zero and draws a steep curve like an exponential distribution to the left and right.

That is, the frequency of occurrence of d-0 is extremely high, for example, 50.

This shows that the correlation between scanning lines is extremely strong.

Therefore, as shown in Figure 3, D(0) has the shortest 1
Assign bit 0, and if ι≠0, D(±ι)
=S(±)L2(ι)(ι>0), and S(±) is assigned the code shown in Fig. 3, and L2(ι)(ι>0).
2(ι) is assigned the pit-by-pit symbol shown in FIG.

This pit-by-pit code is a code word suitable for encoding variables with an exponential distribution.

Therefore, as shown in Fig. 5, D(+l) is 110ι2(
ι), and D(-ι) becomes 111L2(ι).

Regarding the code word of L(ι), when the previous mode is mode 0 or mode 3, L(ι) = 10L1(ι), and when the previous mode is mode 1 or mode 2, L(ι) = 10L2( ι).

The specific code assignment of this L1(ι) is shown in FIG.

Such code assignment enables efficient encoding when the distribution of ι forms a geometric distribution.

The distribution of run lengths in mode 0 or mode 3 is close to the geometric distribution, so if the reference coordinate is P2 and the previous mode is mode 0 or mode 1, ι is exactly equal to the run length of that mode. , ι can be efficiently encoded by using the code L1(ι).

FIG. 7 shows an example of a mode transition code, m1
is the previous mode, and m2 is the current mode.

Furthermore, the length of the mode transition code is assigned according to the mode transition probability.

Generally, the probability of transitioning from mode m1 (m1≠0) to mode 0 is 50% or more, which is extremely higher than the probability of transitioning to other modes.

Therefore, the shortest code 0 is assigned to T(m1/0), (m1≠0).

Such code assignment allows extremely efficient encoding.

Figure 8 shows an example in which the second scanning line A2 and third scanning line A3 in Figure 1 are encoded using the first scanning line A1 as the reference scanning line.The first mode change is 3 bits from the left. The relative coordinate is d=0, and the mode transition is from mode 0 to mode 3.

At this time, the code word is D(0)T(0
/3), and when it is further binarized, it is encoded with only 2 bits as 00 as shown in A4.

The next mode change is the 4th bit, and at this time, d=-
1, ι=1, and the absolute values of d and ι are equal.

The actual binary code length is D(-1)=1110,L(1)
=10000, and this code length is shorter for D(-1).

Therefore, as the relative distance, d=-1 is selected rather than ι=1 and encoded.

The code word is D(-1)T(3/0) as shown in A5, and the binary code is 11100 as shown in A4.

Hereinafter, mode changes occur at the 9th, 10th, and 14th bits.

Furthermore, in order to indicate the end of encoding of the second and third scanning lines, it is assumed that a mode change has occurred from the 16th bit to the virtual 17th bit at the last position.

Here, this mode change is regarded as mode 0 to mode 1.

The code word for this part is D(-3)T as shown in A5.
(0/1)L(1)T(1/3)D(+1)T(3/0
)D(0)T(0/3), and the binary code becomes as shown in A4.

FIG. 9 shows an embodiment of the encoder according to the present invention, where 10 is a signal input line of the second scanning line A2, 11 is a signal input line A3 of the third scanning line, and both signals are binary. It is assumed that the data is digitized and sampled by a clock.

The signals of the second and third scanning lines are sent to the mode change detector 12 in parallel.
The mode change detector 12 checks whether there is a change between the previous mode and the current mode, and if there is a change, the P1 detection circuit 13 detects the coordinates of this change position P1.

This P1 detection circuit 13 simultaneously detects the direction of change in P1.

Note that the signals of the third scanning line are stored in the line memory 14 at the same time.

A line memory 15 stores the signal of the first scanning line 1.

When the P1 detector is activated, the P5 detection circuit 16 reads the first scanning line signal from the line memory 15, detects the level change positions P3 and P4, and also detects the change direction at P3 and P4.

Furthermore, among the changing directions of P3 and P4, the changing position having the same changing direction as P1 is set as P5, and the detection of P5 is completed.

The P1 detection circuit 13 stores the immediately previous position of P1 as the value of P2, and the comparator 17 stores d=P1-P5 and ι=
Calculate P2-P5, compare the absolute values of d and 2, use the smaller absolute value as the reference coordinate, and use the relative coordinate encoder 18 to set P1.
and the relative position of the reference coordinates, and sends an encoded signal 20 to the transmission path through the code configuration circuit 19.

If the absolute values of d and ι are equal, the relative coordinate encoder 18 encodes d and ι together, and the one with the shorter code length is adopted as the code sent to the transmission path.

After encoding the relative position, the mode transition code generation circuit 21
A mode transition code is generated and sent to the transmission path through the code configuration circuit 19.

The above process is repeated until the change point reaches the last coordinate on the scanning line, and when the encoding of the second and third scanning lines is completed, the contents of the line memory 14 are transferred to the line memory 15, and the P1 detection circuit 13 The current value of P1 is stored as P2, and batch encoding of two new scanning lines is started.

As explained above, according to the present invention, the level change position on the first scanning line, which has already been encoded, and the level change position on the second and third scanning line
From any of the previous mode change positions on the scanning line, select the change position closest to the current mode change position on the second and third scan lines as the reference coordinates, and calculate the relative distance between this reference coordinate and the current mode change position. By repeating the encoding of the current mode change position as , and collectively encoding and transmitting the second and third scanning lines, it is possible to make maximum use of the two-dimensional correlation of the image signal. This makes it possible to significantly suppress redundancy, reduce the amount of coded data to be transmitted, and compress the required transmission time or bandwidth. If used as pre-processing for image signal storage, the present invention has the advantage of significantly reducing the required memory capacity, so the present invention can provide a highly effective binary signal encoding and transmission method. .

[Brief explanation of drawings]

FIG. 1 is a diagram showing an example of pixel information on a scanning line to be encoded, FIG. 2 is an explanatory diagram of mode transition, and FIGS. 3 to 5 are diagrams showing relative code words. FIG. 6 is a diagram illustrating mode transition, FIG. 7 is a diagram illustrating an example of mode transition codewords, and FIG. 8 is a diagram illustrating encoding. 9 is a diagram showing an embodiment of the encoder of the present invention. 10, 11...Signal input line, 12...
Mode change detector, 13...P1 detection circuit, 1
4, 15... Line memory, 16...
P5 detection circuit, 17... Comparator, 18...
...Relative coordinate encoder, 19... Code configuration circuit,
20...Encoded signal, 21...Relative coordinate encoder.

Claims (1)

[Scope of Claims] 1. In a method of encoding and transmitting pixel information of second and third scanning lines immediately after the first scanning line based on a first scanning line that has already been encoded and transmitted, , in order to encode and transmit the first position where the mode changes, which is a collection of one pixel each on the top and bottom on the third scanning line, and the position where the mode changes on the second and third scanning lines, and a second position that changes immediately before the first position; and a third position that is a position on the first scanning line where the pixel level changes, is after the second position in the scanning direction, and changes first. and a fourth position on the first scanning line where the pixel level changes, which is after the third position in the scanning direction and where the pixel level changes first; A binary signal code characterized in that one reference coordinate is selected from a fourth position based on a predetermined rule, and the relative position between this reference coordinate and the first position is encoded and transmitted. transmission method. 2. In a method of encoding and transmitting pixel information of the second and third scanning lines immediately after the first scanning line based on the first scanning line that has already been encoded and transmitted, The first mode in which the mode consisting of a collection of one pixel each above and below changes.
a second position that changes immediately before the first position, and a second position that changes immediately before the first position; a third position at a position where the pixel level changes and which is after the second position in the scanning direction and changes first; and a position on the first scanning line where the pixel level changes and the third position a fourth position that is located after the position in the scanning direction and changes first, and if there is a change in level in the scanning direction at the first position on the second or third scanning line, Either one of the changing directions on the second and third scanning lines is set as the changing direction of the first position, or the second
, when there is a change in any of the third scanning lines, the direction of change on the second scanning line is the direction of change in the first position, and the change occurs in the third position or the fourth position, and Among the level change directions in the scanning direction on the first scanning line, a position having the same level change direction as the level change direction at the first position is defined as a fifth position, and the second position and the fifth position
Among the positions, the position with the shortest absolute value among the relative coordinates with the first position is set as the reference coordinate, and if the absolute values of the relative coordinates are equal, the code length will be shortened during encoding. A binary signal encoding transmission system, characterized in that the first position is set as a reference coordinate, and the relative position between the reference coordinate and the first position is encoded and transmitted.