JP3189068B2 - Image data encoding - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a coding method for image data, and more particularly to a coding method for image data in which pseudo halftone binary image data such as photographs and simple binary image data such as characters are mixed. About.

[0002]

2. Description of the Related Art In a conventional image data encoding method, when a halftone image such as a photograph is converted into binarized image data, the image data of n pixels is binarized by binarizing it with an n × n dither matrix. Has periodicity. Figure 7 using this
As shown in (1), exclusive OR (ExOR) is repeated for n input pixels sequentially to convert image data, thereby reducing data change points. This is called convolution processing. As described above, by reducing the number of change points of the image data, the data compression rate at the time of encoding is increased.

[0003]

In the above-described conventional image data encoding method, when a fixed encoding process is performed on binarized data of a photograph / character mixed image, simple binary digitization of characters is performed. Since the processing with high data compression efficiency at the time of encoding is performed on only one of the data part and the pseudo halftone data part of the photograph, the data compression ratio of either the character data part or the photograph data part increases. However, the other has a problem that the data compression ratio is lowered, and the data compression ratio is lowered as a whole.

[0004]

According to the present invention, the character 2
Each of the transmitter and the receiver for transmitting image data in which the binarized image data and the pseudo halftone binary data are mixed,
Dividing the image data for each line into arbitrary blocks,
Character 2 for performing pre-encoding processing on each image block
It has two prediction paths for a binarized image and a pseudo-halftone binarization, and the prediction path for the character binarized image includes a peripheral region using a predicted reference pixel arrangement for the character binarized image data. A pseudo-intermediate image predictor for character binary image data for predicting a current pixel from a pixel state, and a first correctness / error determination unit for comparing the prediction result of the prediction unit with image data to determine correctness / error. A prediction path for halftone binarization includes a pseudo halftone data prediction unit that predicts the current pixel from the state of surrounding pixels using a prediction reference pixel for pseudo halftone binary image data, A second correctness / incorrectness determination unit that compares the result with the image data to determine correctness / incorrectness , and applies a prediction path for each block to a line of a block to be encoded.
On the other hand , encoding and decoding of image data by adaptively and interlocking switching between a transmitter side and a receiver side by validating a prediction path with few errors among the two correct / false judgment results in a nearest block of the previous line and validating the path. And a coding method of image data characterized by performing the following.

[0005] The right / wrong judgment unit can be constituted by an EXOR circuit.

Further, there is provided an image data coding method, characterized in that a smoothing section is provided before the coding section of the transmitter and an inverse smoothing section is provided after the decoding section of the receiver. be able to.

[0007]

Next, the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of one embodiment of the present invention.

The scanner unit 2 of the transmitter unit 1 sends image data obtained by reading an original such as a photograph including a halftone image and characters mixed therein to the A / D conversion unit 3 and converts the image data into multi-valued image data. 4 to generate binary image data in which pseudo halftone binary image data (photo data) and simple binary image data (character data) are mixed.

[0010] The binarized image data is sent to a selective coding pre-processing unit 5 for each block to make it possible to efficiently compress the data at the time of encoding, and then passed to an encoding unit 6. Encoded. The coded data is sent to the receiver unit 7 and decoded by the decoding unit 8, and the image data restoration unit 9 corresponding to the block-specific selective pre-processing unit 5
Is returned to binary image data, and sent to the printer unit 11 through the memory buffer unit 10.

FIG. 2 is a first detailed block diagram of the block-specific selective pre-processing unit 5 and the image data restoration unit 9 according to the present embodiment. First, a block dividing unit (hereinafter abbreviated as a dividing unit) 13 on the transmitter side 12 divides one line of image data into a predetermined number (n) of blocks. The block-divided data is sent to the selector 17 in parallel by a convolution processing unit (hereinafter abbreviated as a processing unit) 14 and two paths. A change point number comparison unit (hereinafter abbreviated as a comparison unit) 15 compares the change point number of the output data of the processing unit 14 detected by the counter 101 with the change point number of the output data of the division unit 13 detected by the counter 102 for each block. , The number of change points is smaller, and the information is notified to an information storage unit (hereinafter abbreviated as storage unit) 16.

In the storage unit 16, n initial values for the first line are set in advance, and in the case of the first line, the output data of the dividing unit 13 is output from the processing unit 14 for all blocks. The selector 17 is informed to select only one of the data, and from the next line, in the previous line, the one having the smaller number of change points detected by the comparator 15 stored in each block is notified to the selector 17 for each block. . The selector unit 17 selects the data notified from the storage unit 16 for each block, and outputs the data to the encoding unit 18.
8 encodes and outputs the received data. Note that a reset signal is input from the dividing unit 13 to the processing unit 14 and the counters 101 and 102 at each block start.

The decoding section 20 on the receiver side decodes the received data. The decrypted data is supplied to the selector 2
2, a deconvolution processing unit (hereinafter abbreviated as an inverse processing unit) 21
Are sent in parallel through two paths, a path sent after being subjected to reverse convolution processing and a path sent as it is. The selector unit 22 outputs data output from the inverse processing unit 21,
Among the data output from the decoding unit 20, the one notified from the information storage unit 25 is output for each block. The information storage unit 25 has the same first one as the information storage unit 16 in advance.
N initial values for the lines are set, and in the case of the first line, whether the output data of the inverse processing unit 21 is the output data of the decoding unit 20 The selector unit 22 is informed to select only the か.

The change point number comparison unit 24 includes a counter 104
The number of change points of the output data of the selector unit 22 detected in
The output data of the selector unit 22 detected by the counter 103 is subjected to convolution processing by the convolution processing unit 23. The number of change points of the data is compared for each block, and the smaller number of change points is notified to the information processing unit 25. The information processing unit 25 stores the information for each block and from the second line to the selector unit 22. For each block, if the number of change points of the data output from the processing unit 23 is smaller, the inverse processing unit 21 Is notified to select the data from the decoding unit 20 if the number of change points of the data output from the selector unit 22 is smaller. Note that a reset signal is input from the decoding unit 20 to the inverse processing unit 21, the processing unit 23, and the counters 103 and 104 at each block start.

FIG. 3 is a second detailed block diagram of the block-specific selective pre-processing unit 5 and the image data restoring unit 9 of this embodiment. In FIG. 3, a prediction method is adopted as a data compression method. The prediction method uses the fact that the original pixel to be predicted has a strong correlation with an adjacent pixel and a pixel separated by the cycle of the input pixel, and predicts the original pixel from the state of peripheral pixels and compares it with the input data. Thus, the image data is converted into the determination data to improve the coding efficiency.

FIG. 5 shows the arrangement of reference pixels at the time of prediction from 10 peripheral pixels. Reference numeral 48 denotes a reference pixel arrangement for prediction with respect to binarized image data such as characters, and reference numeral 49 denotes a reference pixel arrangement for prediction with respect to pseudo halftone binary data such as a photograph, which has a 4-pixel cycle pattern.

In FIG. 3, a block dividing section 26 divides one line of input image data into a predetermined number (n) of blocks, and each of a character predicting section 27 and a photograph predicting section 29. Error determination unit (ExO
R) Output to 28 and 30. In the prediction units 27 and 29, 4
The original pixel is predicted from the reference pixel arrangement of 8,49. ExO
In R28, the prediction result of the prediction unit 27 and the input data
The xOR 30 performs an exclusive logical operation on the prediction result of the prediction unit 29 and the input data to output error determination data for the prediction for each block. ExOR2
The error determination data output from 8 and 30 is sent to the selector 33 in parallel. The change point number comparison unit 31 compares the change point number of the output data of the ExOR 28 detected by the counter 105 with the change point number of the output data of the ExOR 30 detected by the counter 106 for each block, and determines which data has a smaller change point number. And informs the information storage unit 32 of the information for each block.

In the information storage section 32, n initial values for the first one line are set in advance, and in the case of the first line, the output data of the ExOR 28 or the output data of the ExOR 30 is used for all the blocks. The selector 33 is notified so as to select only one, and from the next line, the selector 33 is notified of the change point detected by the comparing unit 31 which is stored for each block at the time of the previous line and is smaller. The selector unit 33 selects and outputs the data notified from the storage unit 32 for each block. The smoothing unit 34 smoothes the error determination data output from the selector unit 33 and passes it to the encoding unit 35, which encodes the data and outputs the encoded data. Note that a reset signal is input to the counters 105 and 106 from the block dividing unit 26 at the beginning of each block.

In FIG. 4, a decoding section 36 decodes and outputs the received data, and an inverse smoothing section 37 performs inverse smoothing of the data, and outputs error determination data of original pixels to an image data restoring section 38. Output. At the same time, the restoration unit 3
The image data (up to the previous pixel) restored and output by the block 8 is, for each block, selected by the selector section 41 from the character prediction section 39 or the photograph prediction section 40 notified from the information storage section 47.
, The predicted original pixel is input to the restoration unit 38 together with the error determination data. And restoration unit 3
At 8, original pixels are specified, and raw image data is output from the restoration unit 38.

In the information storage unit 47, n initial values for the first one line, which are the same as those in the storage unit 32 of FIG. 3, are set in advance. The output of the restoring unit 38 is output to the prediction unit 39 or 40 for all blocks.
The selector section 41 is instructed to output to only one of these.

The character prediction unit 42, the photograph prediction unit 44, and the respective error determination units (ExOR) 43 and 45 use the image data output from the restoration unit 38 in the same manner as the transmitter side and the transmitter side. Convert to judgment data and output. The change point number comparing unit 46 detects the ExOR detected by the counter 107.
The number of change points of the output data of the ExOR 43 is compared with the number of change points of the output data of the ExOR 45 detected by the counter 108 for each block, and the information storage unit 47 is notified of the smaller number of change points. The information storage unit 47 stores the information for each block, and from the second line to the selector unit 41, E
If the change point of the data output from the xOR 43 is smaller, the prediction unit 39 is notified to send the data from the restoration unit 38 to the prediction unit 40 if the change point of the data output from the ExOR 45 is smaller. Note that the counter 10
A reset signal is input to the decoding unit 36 at the beginning of each block.

[0021]

As described above, according to the present invention, input image data is divided into several blocks, and adaptive block-by-block processing is performed on a block-by-block basis based on coding pre-processing selection information of the nearest block of the previous line of the image. By switching the pre-encoding process to the other, there is an effect that the entire image data in which characters and photos are mixed can be efficiently encoded and the data can be compressed.

Further, since the pre-coding process is selected in cooperation with the transmitter and the receiver, an additional pre-coding information for notifying the pre-coding selection information for each block from the transmitter to the receiver is added. Another effect is that a memory buffer for storing the information and the additional information becomes unnecessary.

[Brief description of the drawings]

FIG. 1 is a block diagram of one embodiment of the present invention.

FIG. 2 is a detailed block diagram of a part of FIG.

FIG. 3 is a detailed block diagram of a part of FIG. 1;

FIG. 4 is a detailed block diagram of a part of FIG. 1;

FIG. 5 is a diagram for explaining the operation of the present embodiment.

FIG. 6 is a diagram for explaining the operation of the present embodiment.

FIG. 7 is a diagram for explaining the operation of the present embodiment.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Transmitter part 2 Scanner part 3 A / D conversion part 4 Binarization processing part 5 Selective coding preprocessing part for every block 6 Encoding part 7 Receiver part 8 Decoding part 9 Image data restoration part 10 Memory buffer part 11 Printer section

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-6-30280 (JP, A) JP-A-3-274967 (JP, A) JP-A-63-197172 (JP, A) JP-A-55-19772 41080 (JP, A) JP-A-5-260318 (JP, A) JP-A-4-6954 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04N 1 / 41-1 / 419

Claims (3)

(57) [Claims] 1. A transmitter and a receiver for transmitting image data in which character binary image data and pseudo halftone binary data are mixed, the image data for each line is divided into arbitrary blocks,
Character 2 for performing pre-encoding processing on each image block
It has two prediction paths for a binarized image and a pseudo-halftone binarization, and the prediction path for the character binarized image uses a prediction reference pixel arrangement for the character binarized image data to surround the prediction path. A pseudo-intermediate processing unit including a character binarized image data prediction unit that predicts a current pixel from a pixel state, and a first correctness determination unit that performs a correctness determination by comparing a prediction result of the prediction unit with image data. The pseudo-halftone 2
A pseudo-halftone data prediction unit that predicts the current pixel from the state of surrounding pixels using a prediction reference pixel for the coded image data, and a second unit that compares the prediction result of the prediction unit with the image data to determine whether the pixel data is correct or not. Correctness / error judgment unit, and the application of the prediction path for each block is subject to the encoding.
For the line of the block, the prediction path with few errors is validated among the two correct / false judgment results in the nearest block of the previous line, and the transmitter side and the receiver side switch adaptively and interlockingly to code the image data. A coding method of image data characterized by performing coding and decoding. 2. The image data encoding method according to claim 1, wherein said correctness judgment section is constituted by an EXOR circuit. 3. The image data encoding method according to claim 1, wherein a smoothing unit is provided before a coding unit of the transmitter, and an inverse smoothing unit is provided after a decoding unit of the receiver. The image data encoding method according to claim 1 or 2, further comprising: