JPS61212967A - Mr encoding system - Google Patents

Abstract

PURPOSE:To execute encoding at a high speed by converting a signal to an MR code, based on a counting value of until a change point of inputted black and white. CONSTITUTION:A binary signal of a black and white image which has been received through a bus driver receiver 24 is converted to a series signal by a parallel/series converter 22 and applied to a pre-processing part 18. In the pre-processing part 18, a change point of black and white of the binary series signal is detected by a change point detecting part 19 and a length of white or black is counted by a counter 20 and its count contents are applied to a buffer register 21. When the change point is detected, a DMA request is applied to a DMA control part 17 from the detecting part 19, the contents of the register 21 are transferred to a RAM 14 by, for instance, a byte unit, and when the counter 20 reaches a prescribed value, an interrupting signal is applied to a microprocessor 11. By its interruption, the processor 11 executes a MR encoding processing, based on the contents of the RAM 14.

Description

[Detailed Description of the Invention] [Summary] A device that detects the changing point of an input black and white binary signal, identifies the position of the signal changing point based on the count counted up to this changing point, and converts it into an MR code. It is.

[Industrial application field]

The present invention utilizes MR, which is a type of two-dimensional encoding, using a microprocessor to compress and store or transmit a binarized black and white image signal in an image processing communication system, etc. This relates to an MR encoding method for converting to a (ModifiedREAD) code.

M H (Modify
ed Huffman) code or the above-mentioned MR code. The MH encoding method encodes a white or black run length on a scanning line, and corresponds to -dimensional encoding. Further, the MR encoding method performs encoding in accordance with the relative position of black and white change points between the encoded scanning line and the reference scanning line, and is equivalent to two-dimensional encoding. In this case, there is a demand for faster encoding.

[Conventional technology]

In the conventional MR encoding method, an encoding processing unit reads a black-and-white binary signal, determines the point of change of the binary signal from white to black, or from black to white, and then selects the reference scanning line and the encoded scanning line. The code is converted into an MR code by software using three modes: path mode, vertical mode, and horizontal mode, depending on the relative position of the change point in the line. For example, as shown in the CCITT recommendation standards, in Figure 4 (al), the white pixel at the starting point on the encoded scanning line is ao, the black pixel at the first change point is al, and the starting point on the reference scanning line is If the black pixel at the reference change point is b+ and the white pixel at the next reference change point is b2, then the white pixel b2 on the reference scanning line is located ahead of the black pixel al at the change point on the encoding scan line. , b+ on the reference scan line in this case.

The black pixels between b2 and P are not continuous with the black pixels on the encoded scanning line. In this case, it is processed as pass mode. Then, the next encoding is performed using the white pixel ao+ on the encoding scanning line immediately below the white pixel b2 on the reference scanning line as the next starting pixel.

On the other hand, as shown in FIG. 4(b), the black pixel between the black pixel S at the change point on the reference scanning line and the white pixel b2 at the next change point, and the black pixel on the encoding scan line. If they are continuous, they are treated as vertical mode. Also, a black pixel a1 at a change point on the encoded scanning line and a black pixel a1 at a change point on the reference scanning line,
If the number of pixels between the two is 4 or more, it is processed as horizontal mode.

In the encoding process, MR codes are determined depending on the pass mode, vertical mode, and horizontal mode, and FIG. 5 shows the relationship between the pixels to be encoded and the MR codes. Depending on each mode, pixels to be encoded, symbols, MR
The code is determined as shown. MH in horizontal mode in the figure
is Modified Hufman.
ffman) code, and this MH code represents the run length 0 to 63 as a terminating code that determines the switching point between white and black, and when the number of consecutive white or black is an integral multiple of 64, it is represented as a make-up code. It represents. In the vertical mode, whether the black pixel al at the changing point on the encoded scanning line is directly below the black pixel al at the changing point on the reference scanning line with respect to the black pixel al at the changing point on the reference scanning line, or whether it is shifted to the right or left. An MR code is assigned depending on the deviation.

Such conventional encoding processing uses, for example, a processing device such as
It uses a microprocessor and is completely programmed. FIG. 6 shows a processing flow diagram necessary as preprocessing for conventional MR encoding. For example, the binary signal is stored in the memory MMI in units of one scanning line, one byte is read from the beginning of this memory MMI and ill is executed, and it is determined whether the one byte is completely white (2), and whether it is completely black or not. Then, if white and black are mixed in one byte, perform a 1-bit shift (4) sequentially, and use black and white determination (5) to determine the point of change between white and black. , count as bit processing up to the change point 6), add the result 7), store the change point position in memory MM2 as the counted value up to each change point, ao), (al>, (az) (4th 8) and determine whether or not the next process is to be performed (9).
)do. In this way, the white/black change point for one scanning line is identified. In this case, the bit processing part (A) is included.

FIG. 7 is a flowchart of the encoding process, in which change points (b+), (t)z) of the reference scanning line are read out from the memory MM3 (11). This memory MM3 is used to transfer change point position information of the previous encoded scanning line from the memory IJMM2, or to switch the memory MM2 as the memory MM3. Next, from the memory MM2, the change point (ao) of the encoded scanning line obtained in the preprocessing flow shown in FIG. 6, (a
1) and (az) are read out (12), a mode is determined based on the relative relationship between the change points of the reference scanning line and the encoded scanning line, and encoding processing is performed (13). Note that the processing in FIG. 7 is byte processing.

[Problem that the invention seeks to solve]

In the MR encoding method in the conventional image processing communication system, as explained in FIG.
Process (A) in the figure) detects the point of change from white to black or from black to white, and performs MR encoding according to the relative position of the point of change between the reference scanning line and the encoded scanning line. and
It was a completely software-based process.

Therefore, the microprocessor is required to perform bit processing in order to detect a change point, and by using a microprocessor for byte processing for focus processing, there is a drawback that the MR encoding processing time becomes longer.

The present invention improves the drawbacks of the conventional technology, and by adding an M simple configuration, high-speed MR can be achieved by byte processing.
The purpose is to perform encoding.

[Means for solving problems]

The MR encoding system of the present invention will be explained with reference to the principle block diagram of FIG. The change point detection unit includes an encoding processing unit 1, a change point detection unit 3 that detects a change point of two black and white system codes converted into a serial signal, and a counting unit 2 that counts the two system codes. 3, the count value when the black and white change point of the 2 plant code is detected is notified to the encoding processing unit 1, and the encoding processing unit 1,
The black and white changing point position is identified from the counted value and converted into an MR code based on the relationship between the changing point position of the reference scanning line and the changing point position of the encoded scanning line.

[Effect]

Since the counting unit 2 simply counts the two types up to the black and white change point, it can obtain the run length of white or black, and the encoding processing unit 1 reads this run length information in bytes, The software identifies the change point position,
The code changes to an MR code depending on the relationship with the pixel to be coded shown in FIG. That is, since encoding can be performed only by hide processing, which is not shown in FIG. 7, it is possible to increase the speed.

〔Example〕

Embodiments of the present invention will be described in detail below with reference to the drawings.

FIG. 2 is a block diagram of an embodiment of the present invention, and portions corresponding to the configuration in FIG. 1 are indicated by numbers in parentheses. In the figure, 11 is, for example, a microprocessor (MPU),
12 is a clock generator (CLG) for generating a clock signal to be applied to the microprocessor; 13 is a read-only memory (ROM) in which a program is stored; 14
is random access memory (RAM) that stores data.
), 15 is an address bus (AB), 16 is a data bus (DB), 17 is a DMA (direct memory access) control unit (DMAC), 1B is a preprocessing unit, 19 is a change point detection unit (DET) (3 ), 20 is a counter (CT
R) (2) , 21 is a buffer register (BR),
22 is a parallel/serial converter (P/S) for converting a black and white binary signal input as a parallel signal into a serial signal, 23 is a data register (DR), and 24 is a bus driver/receiver (DV) 1.25 is the run length register (RLR).
), 26 is a conversion unit (RLB) that converts the run length into a black and white binary image signal, and 27 is a parallel-to-serial converter (S/P) that converts the image signal as a serial signal into a parallel signal to be sent to the lot. ), 28 are buses for connecting with other microprocessors, etc.

The 2(i) signal of the black and white image received via the bus driver/receiver 24 is a parallel signal with the number of bits corresponding to the width of the bus 28, so it is converted into a serial signal by the parallel-to-serial converter 22 and then processed by the preprocessor. 18. This preprocessing section 18 is composed of a change point detection section 19, a counter 20, and a buffer register 21. Detected, the length of white or black is counted by the counter 20, and the count contents are added to the buffer register 21.

When the change point detection is performed, the change point detection unit 19 sends the DMA
A DMA request is applied to the control unit 17, and the DMA control unit 17
is buffer register 2 indicating white or black run length
Control is performed to transfer the contents of 1 to the random access memory 14 by DMA transfer, for example, in byte units, and when the counter 20 counts up to a predetermined value, an interrupt signal is applied to the microprocessor 11. Upon interrupt, the microprocessor 11 reads the contents of the random access memory 14, identifies the position of the change point from the count indicating the white or black run length, and determines the relative position of the change point between the reference scan line and the encoded scan line. Depending on the position, processing is performed to convert it into an MR code.

The microprocessor 11 has a random access memory 1
4, white or black run length information is read in byte units, and therefore it is possible to identify black and white change points by byte processing. Also, the converted MR code signal is stored in the random access memory 14 or added to the data register 23 and sent to the bus driver/receiver 2.
4 to the bus 28.

In addition, when transferring image information consisting of an MR code signal to a display device, facsimile device, etc., the MR code signal is converted into a run-length signal under the control of the microprocessor 11, added to the run-length register 25, and converted according to the run-length signal. The unit 26 converts it into a black and white serial binary signal, and the serial/parallel converter 27 converts it into a parallel binary signal with the number of bits corresponding to the width of the lotus 28.
The receiver 24 sends out a black and white two-print number to the bus 28.

FIG. 3 is a block diagram of main parts of an embodiment of the present invention, showing a preprocessing section and a parallel-to-serial converter. In the figure, 31 is a flip-flop, 32 is a buffer register (BR), 33 is a parallel-to-serial converter (P/S),
34 is a frequency divider that divides the clock signal into 1/8, 35 is a change point detection section, 36 is an exclusive OR circuit, 37 and 38 are flip-flops, and 39 is an AND circuit that controls supply and stop of the clock signal. , 40 is a counter, 41 is a buffer register (BR), 42 is an AND circuit for detecting that the count content of the counter 40 has reached a predetermined value, 43
is a flip-flop.

The frequency division signal from the frequency divider 34 is applied to the clock terminal C of the flip-flop 31, and "1" is always added to the data terminal, so when the frequency division signal is applied, the Q terminal becomes "1". Then, the DMA request signal DRQ is output, and the D
It is sent to the MA control unit. Accordingly, when the DMA enable signal DOK is applied from the DMA control section, the flip-flop 31 is reset and the buffer register 32 sets the monochrome flip value signal.

Further, the flip-flop 37 of the change point detection section 35 is reset when the permission signal OK from the microprocessor becomes "1", and the signal applied from its d terminal to the AND circuit 39 becomes "1", so that the clock signal CLK is applied to a parallel to serial converter 33, a frequency divider 34 and a counter 40.

The parallel-to-serial converter 33 sets two numbers in parallel from the buffer register 32 according to the frequency division signal, outputs it as a serial signal according to the clock signal, and adds it to the exclusive OR circuit 36 of the change point detection section 35. In other words, the two 8-bit parallel sets are sequentially shifted in accordance with the clock signal and converted into a serial signal.

The output signal of the exclusive OR circuit 36 is output from the flip-flop 3
7.38 is applied to the clock terminal C of the counter 40, the data terminal of the counter 40, and the buffer register 41, and when the output signal becomes "0", the flip-flop 37.38 performs an inversion operation, and the counter 40 is applied to the buffer register 41. The count content of is set, and the counter 40 is cleared by the clock signal CLK.Therefore, a serial signal is applied to the exclusive OR circuit 36, and for example, if "1" indicating black is continuous, the counter 40 is cleared by the clock signal CLK. Then, the Q terminal of the flip-flop 38 becomes 0'', and the counter 40 counts the clock signal CLK via the AND circuit 39. This corresponds to counting the number of black pixels, and the count indicates the run length of black pixels.

Next, when the serial signal changes to white in the middle and the serial signal reaches "0°", the output signal of the exclusive OR circuit 36 becomes "0".Thereby, the Fritsop flops 37 and 38 perform an inverting operation. , and the buffer register 41 sets the count contents of the counter 40 at that time. Also, the Q terminal of the flip-flop 37 becomes "bul", and the d terminal becomes "
0''L, the clock signal CLK supplied via the AND circuit 39 is stopped, and the operations of the frequency divider 34, parallel-to-serial converter 33, and counter 40 are stopped. A request signal RQ is sent to the microprocessor, data (count contents) is transferred from the buffer register 41 in response to a permission signal OK from the microprocessor, and the flip-flop 37 is reset.

When the flip-flop 37 is reset, the clock signal CLK is supplied again via the AND circuit 39, and since the output signal of the exclusive OR circuit 36 is "'O" at that point, the counter 40 is cleared, and then Parallel-serial converter 3
When a signal of "0" indicating white, which was stopped in the middle of conversion, is output from 3, the Q terminal of flip-flop 38 becomes 1.
Since the output signal of the exclusive OR circuit 36 becomes 1", the counter 40 counts the clock signal. In other words, the number of white pixels is counted, and the count content is 1". This indicates the run length of the pixel.

Also, as a reminder, when the same black pixels continue and the counter 40 reaches a predetermined count, for example 255, the output signal of the AND circuit 42 becomes "1" and is applied to the clock terminal C of the flip-flop 43. Therefore, the Q terminal of the flip-flop 43 becomes "1" and becomes an interrupt signal IR to the microprocessor.When the permission signal OK is applied from the microprocessor, the flip-flops 37 and 43 are reset.

Run length information of white and black pixels is added to the microprocessor, and the microprocessor is able to identify the position of a change point from the starting position of a scanning line through byte processing.

Since the change point positions can be easily identified for each of the reference scanning line and the encoded scanning line, the conversion process to the MR code can be performed according to the relationship shown in FIG. 4.

〔Effect of the invention〕

As explained above, according to the present invention, the encoding processing unit 1
It is no longer necessary to detect the change point between white and black in the image by bit processing, but it can be done by byte processing. It has the advantage of being faster.

[Brief explanation of drawings]

FIG. 1 is a block diagram of the principle of the present invention, FIG. 2 is a block diagram of an embodiment of the present invention, and FIG. 3 is a block diagram of main parts of the present invention.
Fig. 4 (81, (bl is an explanatory diagram of pixels on a reference scanning line and an encoding scanning line, Fig. 5 is an explanatory diagram of encoding target pixels and MR code, Fig. 6 is a conventional encoding preprocessing Flowchart, FIG. 7 is a flowchart of encoding processing. 1 is an encoding processing section, 2 is a counting section, 3 is a change point detection section, 1
1 is the microprocessor (MPU), 20.40 is the counter (CTR), 1.19.35 is the change point detection unit (DE
T), 13 is read-only memory (ROM), 1
4 is random access memory (RAM), 17 is DM
A control unit, 18 is a preprocessing unit, 21.41 is a buffer register (BR), 22.33 is a parallel-to-serial converter (P/S)
It is.

Claims (1)

[Claims] In an MR encoding method that converts into an MR code based on the relative position of a change point between a reference scanning line and an encoded scanning line, a counter that counts black and white binary signals converted into a serial signal. a change point detection section (3) for detecting a change point of the binary signal; and an encoding processing section (1), wherein the change point detection section (3) detects a change point of the binary signal. When a black and white change point is detected, the value counted by the counting section (2) is given to the encoding processing section (1), and the encoding processing section (1) determines the black and white transition position based on the counted value. An MR encoding method characterized by identifying and encoding.