JPH0435363A - Coding method for binary picture - Google Patents

Abstract

PURPOSE:To effectively compress a data of a binary picture by processing a binary picture into a data with a prescribed command code and a run length code for each of two main scanning lines to make a data identifying the combination of picture element sets to be one bit respectively. CONSTITUTION:A memory 3 connects to a data input section 2, and a block level picture element detection circuit 4 and a white level pictrue element detection circuit 5 and a different color level picture element set detection circuit 6 are connected in parallel with the memory 3. That is, a binary picture resulting binarizing picture information in a matrix shape is subject to read scanning for each of two main scanning lines, consecutive number in the main scanning direction is detected, coded by a run length code preset for each color, a set of picture elements each of which is arranged in the order of white and black levels in the subscanning direction and a set of picture elements each of which is arranged in the order of black and white levels in the subscanning direction are detected respectively to set a 1-bit identification data, and number of consecution in the main scanning direction is detected for the set of picture elements of black and white levels and white and black levels to which the one-bit identification data is set and they are processed into data. Thus, the data is compressed in an excellent efficiency.

Description

[Detailed description of the invention]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a binary image encoding method used for compressing data transmitted by a facsimile machine. Conventional technology In facsimile machines and the like, binary images to be transmitted are encoded and compressed in order to reduce line usage time when transmitting image information. The method is MW (Modified Hu
ffn+an) method. This is an encoding method that has been standardized in Group 3 facsimile machines, and a Huffman code is set based on the frequency of appearance statistically determined in advance for black and white run lengths. Encoding allows the amount of data in binary images such as documents to be compressed. For example, the number of dots on the main scanning line of an A4 sheet of paper is 1728, but extremely long run lengths rarely appear, so as shown in Table 1.2 below, in the MH method, the number of dots on the main scanning line is 63. A terminal code is set only for the following run lengths, and an assembly code is added to the terminal code for run lengths of length 64 or more. Table 1 Table 2 Problems to be Solved by the Invention By using the above-mentioned MH encoding method, it is possible to compress the data amount of binary images such as documents, but this mainly applies to binary image data. Since this is a one-dimensional method that compresses only in the scanning direction and not in the sub-scanning direction, there is a need for a binary image encoding method that can compress data more effectively. For example, there is a two-dimensional MR (Modified Read) method that compresses data more effectively than the MH method by reflecting the data of the Moe line when compressing data on the - line, but this method requires an extremely difficult processing method. The productivity of the device is low because it requires complicated and dedicated circuit parts. Means for Solving the Problem A binary image obtained by converting image information into a matrix is read and scanned every two main scanning lines, and each of a pair of pixel sets is scanned in the sub-scanning direction of the scanned main scanning line. It detects whether the pixels are both white, black, or different colors, and when the same color is detected, a same color indicator is set at the beginning, and the number of consecutive pixels in the main scanning direction is detected, and each color is Each pixel set is encoded using a preset run-length code, and a pixel set in which a different color is detected is set with a different color indicator code at the beginning, and is divided into a pixel set in which each pixel is black and white in the sub-scanning direction and a pixel set in which each pixel is black and white. 1-bit identification data is set by detecting each pixel pair, and for each black-white and black-and-white pixel set to which this 1-bit identification data is set, the number of consecutive pixels in the main scanning direction is detected and converted into data. . By converting a binary image into data using a predetermined instruction code and run-length code for each two main scanning lines, the binary image can be compressed in the sub-scanning direction as well. Since the data for identifying each image is 1 bit, the binary image data can be effectively compressed by simple data processing. Embodiment An embodiment of the present invention will be described based on the drawings. Note that a facsimile apparatus 1 will be exemplified here as an example of the binary image encoding method. As illustrated in FIG. 2, this facsimile device 1 has a memory 3 connected to a data input section 2 connected to an image reading section (not shown) that binarizes image information into a matrix. A black pixel group detection circuit 4, a white pixel group detection circuit 5, and a different color pixel group detection circuit 6 are connected in parallel to this memory 3. The black pixel group detection circuit 4 and the white pixel group detection circuit 5 are commonly connected to the encoding circuit 8 via the run length counter circuit 7, and the different color pixel group detection circuit 6 is connected to the circuit 7.
A circuit 9 having substantially the same structure as 8 is connected to 10 and a flag adding circuit 11 . Therefore, this flag adding circuit 11 and the encoding circuit 8,
10 are connected via one signal synthesis circuit 12 to a data output section 13 connected to a data transmitter (not shown) or the like. In the facsimile apparatus 1 of this embodiment, the aforementioned Huffman code of the MH system is set as the run-length code in each of the encoding circuits 8 and 10. In this configuration, in this facsimile device 1, first, as illustrated in FIG. 1(a), when binary image data is transferred from the data input unit 2 to the memory 3, ), this binary image data is read and scanned every two main scanning lines by three pixel set detection circuits 4 to 6, and each pixel of a pair of pixel sets is white in the sub-scanning direction. It is detected whether the color is black, black, or different colors from each other. Therefore, for a pixel set for which the same color is detected by the pixel set detection circuits 4 and 5, a flag O, which is a same color indicating code indicating the same color, is set at the beginning of the pixel set, as illustrated in FIG. 1(C). It is converted into 1-bit data as illustrated in FIGS. 3(a) and 3(b), and the consecutive number in the main scanning direction is detected by the run length counter circuit 7.
is converted to Huffman code. In addition, a pixel set in which a different color is detected by the different color pixel set detection circuit 6 is set with a flag l, which is a different color indicating code indicating a different color, at the beginning, and a pixel set in which each pixel is black and white in the sub-scanning direction. Black and white pixel sets are detected and 1-bit identification data of [O] or [1] is set for each pixel set, as shown in Fig. 3 (C
) and (d), and the consecutive number in the main scanning direction is detected by the run length counter circuit 9. Therefore, this data is transferred to the encoding circuit 10.
and converts the black-and-white and black-and-white pixel sets into Huffman codes for white and black in the MH system, respectively, according to 1-bit identification data. Therefore, the data of the pixel set that has been added with various flags and converted into Huffman codes as described above is sent to the signal synthesis circuit 1.
In step 2, the data is combined into a series of data and transferred from the data output unit 13 to a data transmitting device or the like. That is, in the binary image encoding method of this facsimile device 1, the binary image is scanned and encoded every two main scanning lines, and at this time, data identifying the combination of pixel sets is Since it is a bit, data can be effectively compressed with simple data processing. For example, when this facsimile device 1 encodes image information as shown in FIG.

[0] is set, and since this pixel set is white and the number of consecutive pixels in the main scanning direction is 5, it is converted to a white Huffman code [1100] with a run length of 5. Next, a different color pixel set is detected and flag [1] is set,
This is a set of black and white pixels with a consecutive number of 3 and a set of black and white pixels with a consecutive number of 2, so the Huffman code for white with a run length of 3 [10
001 and Huffman code for black with run length 2 [11
] is converted to Next, a same color pixel set is detected and a flag [0] is set, and since this pixel set is a black pixel set with a consecutive number of 2, a Huffman code for black with a run length of 2 [1
1]. Therefore, data (2) of image information encoded as described above and data (2) encoded by the conventional MH method are illustrated below. ■・・・ 011001100011011■・
... 100110i1 As is obvious from the data length described above, the encoding method of this facsimile device 1 can effectively compress binary image data. Note that the encoding method of the facsimile device 1 of this embodiment is as follows.
Since it can be implemented by improving the existing MH type processing circuit in terms of software, the structure is simpler and the productivity is better than that of the MR type processing circuit. Further, in the facsimile device 1 of the present embodiment, the Huffman code of the MH method is used to convert different color pixel sets into data, but the present invention is not limited to the above method. As data conversion, a 1-bit identification flag is set for black and white and black and white pixel sets,
After this, it is also possible to enter a code indicating a preset run length. In this case, data can be expected to be compressed with better efficiency by statistically determining the frequency of appearance of black-and-white and black-and-white pixel sets in advance and setting the code for the run length. Effects of the Invention As described above, the present invention reads and scans a binary image obtained by binarizing image information in a matrix every two main scanning lines,
It is detected whether each pixel of a pair of pixel sets is white, black, or different colors from each other in the sub-scanning direction of the scanned main scanning line, and if the same color is detected, the same color is added to the top of the pixel set. At the same time as setting an instruction code, the number of consecutive pixels in the main scanning direction is detected and encoded using a preset run-length code for each color, and for pixel sets in which a different color is detected, a different color instruction code is set at the beginning and the sub-color is encoded. In the scanning direction, pixel groups in which each pixel is black and white and pixel groups in which each pixel is black and white are detected and 1-bit identification data is set, and the black-and-white and black-and-white pixel groups to which this 1-bit identification data is set are By detecting the consecutive numbers in each main scanning direction and converting them into data, the binary image is also compressed in the sub-scanning direction, and each data identifying the combination of pixel sets is Because it is a bit, binary image data can be effectively compressed with simple data processing, and it has the effect of contributing to increasing the transmission speed of image information in facsimile machines. .

[Brief explanation of the drawing]

FIG. 1 is a conceptual explanatory diagram of data processing showing an embodiment of the present invention, FIG. 2 is a block diagram, and FIGS. 3 and 4 are conceptual explanatory diagrams of data processing.

Claims (1)

[Claims] A binary image in which image information is binarized in a matrix is read and scanned every two main scanning lines, and each pixel of a pair of pixels is white in the sub-scanning direction of the scanned main scanning line. It detects whether the pixels are black or have different colors, and when the same color is detected, a same color indicator is set at the beginning, and the number of consecutive pixels in the main scanning direction is detected and a preset run is set for each color. Pixel sets in which a different color is detected are encoded using a length code, and a different color indicator code is set at the beginning of the pixel set, and pixel sets in which each pixel is black and white in the sub-scanning direction and pixel sets in which each pixel is black and white are detected and 1 bit is generated. 1-bit identification data is set, and each black-and-white pixel set and black-and-white pixel set to which this 1-bit identification data is set is converted into data by detecting the number of consecutive pixels in the main scanning direction. Image encoding method.