JPH0241232B2 - - Google Patents

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention provides an image transmission device such as a facsimile, an image storage device such as an image database, etc., in which binary images and halftone images consisting of characters, line drawings, etc. The present invention relates to an encoding device that compresses data by encoding mixed image signals.

[Prior art]

Conventional facsimile machines and the like mainly consist of devices that handle binary images, and when a halftone image is input into such a device, the statistical properties of the image signal are generally very different from those of a binary image, and the efficiency of encoding is reduced. becomes extremely bad. For this reason, some encoding devices for halftone dot images have been considered, but this device is intended for halftone photographs that display lines, slopes, etc. using a specific screen, and is designed to encode various halftone dot images. It cannot be used for devices that take photographs.

A conventional image signal encoding device will be described below with reference to the drawings. FIG. 1 is a diagram showing an example of reference pixel positions used for predictive encoding in a conventional image signal encoding device. In the figure,
is a coding pixel which is the pixel to be coded, A and I are pixels on the same line as the coding pixel X, and B, C and D are pixels on the line immediately before the coding pixel X. Further, R ₁ is the main scanning direction, and R ₂ is the sub-scanning direction. In predictive encoding, the encoded pixel signal is predicted by the preceding neighboring pixel signal,
A difference signal (prediction error signal) between the predicted value and the actual signal is encoded. FIG. 2 is a table showing an example of a prediction method in the image signal encoding device of FIG. 1. As shown in FIG. 1, the reference pixels in this example include pixel A on the same line as the encoded pixel X and immediately preceding it, and three pixels B, C, and D on the line immediately before the encoded pixel X. It is being
FIG. 3 is a diagram showing an example of a code when a prediction error signal is encoded in the image signal encoding apparatus of FIG. 1. In the figure, m is an integer of 0 or more and 3 or less, n is an integer of 0 or more, and xx is m expressed in 2-bit binary. Fig. 4 is a block diagram showing a conventional image signal encoding device;
FIG. 4 is a diagram showing examples of codes for explaining the operation of the image signal coding device shown in FIG. 4. In each figure, 1 receives an image signal indicating a coded pixel It is a predictor that creates ^. 2 is an exclusive OR (EX-OR) circuit that creates a prediction error signal Y based on the image signal and the prediction signal This is an encoder that performs encoding.

Next, the operation of the conventional image signal encoding apparatus shown in FIG. 4 will be explained. Shown in Figure 5
a _-1 and a ₀ are image signals of n-1 line (line immediately before the encoded line) and n line (encoded line), respectively. One line has 24 pixels,
They are numbered #1 to #24 in scanning order. When scanning the encoded line of a ₀ , at each position, the first
Each of the reference pixels A, B, C, and D shown in the figure looks like c shown in FIG. 5, respectively. Here, if each reference pixel A, B, C, D is outside the effective screen range, it is set to 0, and in FIG. 5, this is shown enclosed in parentheses. The predictor 1 shown in FIG. 4 creates a prediction signal X^ shown as d in FIG. 5 in accordance with the prediction method table shown in FIG. In the exclusive OR circuit 2, this predicted signal X^ and the actual image signal (see Fig. 5)
a ₀ ), a prediction error signal Y shown as e in FIG. 5 is obtained. The encoder 3 performs encoding according to the code example shown in FIG. 3 based on this prediction error signal Y, and eventually obtains the encoded sequence shown by f in FIG. 5. Note that the signal enclosed in parentheses at the end of the prediction error signal Y is virtually added to indicate the end of the line, and the code for this signal is also added to the encoded sequence f. .
In the case of an image signal as shown in Figure 5, the encoding line
After #11 of a ₀ , there is a substantially periodic signal at 4-pixel intervals. (In a halftone image where the horizontal halftone period is 4 pixels apart, many such signals occur.) Therefore, 1
The total code bits of the line are now 31, and the amount of data has increased. This is because prediction for a normal binary image as shown in FIG. 2 is also performed for the halftone image portion, so that the prediction matching rate decreases.

As mentioned above, in conventional image signal encoding devices, even when a normal binary image and a halftone image are mixed, encoding is performed for the normal binary image, so that efficient data processing is possible. Compression was impossible. In addition, since the image signal obtained by reading the halftone dot image was simply encoded into a binary signal, if the halftone dot period and the pixel interval were close, moiré would occur and the image quality would deteriorate significantly. There were drawbacks such as:

[Summary of the invention]

This invention was made with the aim of improving the drawbacks of the conventional ones as described above, and it is possible to calculate the average value of a predetermined area near the pixel for an image signal containing a mixture of binary images and halftone images. The binary image and the halftone image are separated using a low-pass filter that performs processing and a comparison circuit that detects and compares the difference between the maximum and minimum values of the output of the low-pass filter within the area. For point images, the output of the low-pass filter is dithered by a dithering circuit,
Using each predictor for a binary image and a dithered image, each predictor is selectively used to perform predictive conversion on encoded pixels according to the prediction matching rate of a predetermined number of preceding encoded pixels. , an image signal encoding device that can compress data with high efficiency and reproduce moiré-free and high-quality images even for halftone photographs using various screens. This is what we provide.

[Embodiments of the invention]

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

FIG. 6 is a block diagram showing an image signal encoding device according to an embodiment of the present invention, and FIG. 7 shows pixel positions where average value processing is performed in the image signal encoding device of FIG. 6. 8 is a diagram showing a dither threshold matrix in the image signal encoding device of FIG. 6, and FIG. 9 is a diagram showing a prediction method for dithered images in the image signal encoding device of FIG. 6. FIGS. 10 and 11 are diagrams each showing examples of codes for explaining the operation of the image signal encoding device shown in FIG. 6. In FIG. 6, 4 is 3 shown in FIG. 7 for the input image signal.
A low-pass filter 5 that performs average value processing within ×3 pixels counts the difference between the maximum value and minimum value within the area shown in FIG. 7 from the image signal output from the low-pass filter 4. , an area maximum/minimum difference counting/comparison circuit 6 for determining whether the value is 2/9 or more; A dither circuit 7 that performs systematic dithering processing using a threshold matrix performs a binary conversion between the output signal of the dither circuit 6 and the image signal selection signal that is the output signal of the area maximum/minimum difference counting/comparison circuit 5. Binary signal switchers 1a and 1b that select one of the input image signals are for binary images (according to the prediction method shown in FIG. ) and dithered images (based on the prediction method shown in FIG. 9), the predictors 2a and 2b are a binary image predictor 1a and a dithered image predictor 1b, respectively.
8a Exclusive OR circuits for binary images and dithered images that take exclusive ORs of the binary and dithered predicted signals output from the encoded image signal;
and 8b are an exclusive OR circuit 2a for binary images and an exclusive OR circuit 2b for dithered images, respectively.
Prediction matching rate coefficients for binary images and dithered images that count the number of prediction errors within three pixels preceding the encoded pixel X based on the binary and dithered prediction error signals output from A circuit 9 compares the binary and dithered prediction error number signals which are the respective outputs of the prediction matching rate counting circuits 8a and 8b for binary images and dithered images, and calculates the dithered prediction error number signal as a binary image. A prediction matching rate comparison circuit 10 that determines whether the prediction error number signal is larger than the prediction error number signal selects either a binary prediction error signal or a dither prediction error signal based on the prediction error signal selection signal that is the output of the prediction matching rate comparison circuit 9. A prediction error signal selector 3 selects one of the prediction error signals, and 3 is an encoder that encodes the prediction error signal (prediction error signal to be encoded) selected by the prediction error signal selector 10.

Next, the operation of the image signal encoding apparatus shown in FIG. 6, which is an embodiment of the present invention, will be described. a, g, h shown in Figures 10 and 11,
Regarding i, j, k, and l, line numbers are added with subscripts to distinguish them, with the encoded line (n-th line) being 0. Further, regarding d, e, and o, the subscript F is added to the binary one, and the subscript D is added to the dither one. Furthermore, D at i or p
and F represent selection of a dither image signal or a binary image signal or a prediction error signal, respectively. The portion to the right of #11 of the image signal a shown in FIG. 10 includes a halftone dot image with a horizontal halftone dot period and an interval of four pixels. Here, a _-1 and a ₀ are
This is the same as that used in the explanation of the operation of the conventional device shown in FIG. 5 above. For each pixel signal of the image signal shown in the figure, the average value of the nine pixels shown in FIG. 7 is determined by the low-pass filter 4 as its output g. Here, when the pixel for which the average value is to be calculated is outside the effective screen range, the pixel value is set to 0. Next, when the difference between the maximum value and the minimum value within the region shown in FIG. 7 is determined for the output g of the low-pass filter 4, an intra-block difference signal h is obtained. The intra-area maximum-minimum difference counting/comparison circuit 5 obtains an image signal selection signal i based on the intra-block difference signal h (h
dither D when 2/9, binary F when h<2/9).

On the other hand, the output g of the low-pass filter 4 is
Using the dither threshold matrix shown in the figure, a dithered signal k is obtained which is dithered by the dithering circuit 6 (if gj, k is set to 1, otherwise set to 0).
Here, it is assumed that the dither threshold corresponds to 1/8 of the #1 pixel of the n-th line. The dither threshold corresponding to the output g of the low-pass filter 4 is set to the 10th
Indicated by j in the figure. The binary signal switch 7 appropriately selects the input image signal a and the dithered signal k based on the image signal selection signal i to obtain an encoded image signal l. Here, the image signal corresponding to the reference pixel shown in FIG. 1 becomes as shown in FIG. 11c, and in the binary image predictor 1a, according to the prediction method shown in FIG. A binary prediction signal indicated by _F is obtained, and a binary prediction error signal e _F is obtained by the binary image exclusive OR circuit 2a. Further, the dithered image predictor 1b creates a dithered prediction signal shown as _dD in FIG. 11 based on the encoded image signal l according to the prediction method shown in FIG. In the prediction method shown in Fig. 9, not only the reference pixel signal but also the dither threshold value information of the pixel (equivalent to information about the position of the pixel in a 2 x 2 matrix) is taken into account. There is. This is because the properties of dithered images vary greatly depending on the threshold of the pixel, such as whether it tends to be 0 or 1, so taking the threshold into account can improve the prediction matching rate. It is. Further, I is added as a reference pixel because its threshold value is equal to the encoded pixel X, and the correlation is stronger than that of each pixel B or C. The dithered prediction signal _dD is used in the dithered image exclusive OR circuit 2b to obtain the dithered prediction error signal _YeD in the same way as for binary images. The two types of binary and dithered prediction error signals e _F and e _D described above are input to two of the three pixels preceding the encoded pixel value and number of dither prediction errors o _F and
o _D is counted. The prediction error signal selection signal p is obtained by comparing the binary values and the numbers of dither prediction errors o _F and o _D in the prediction matching rate comparison circuit 9 (Here, if o _F > o _D , p is D, otherwise binary F). In the prediction error signal selector 10,
Based on this prediction error signal selection signal p, binary and dither prediction error signals e _F and e _D are selected to obtain a coded prediction error signal e. Finally, encoder 3 gives
Encoding is performed according to the code string shown in FIG. 3 to obtain a coded sequence f. Further, the decoding device decodes the coded sequence f according to the code example shown in FIG. 3 to obtain a coded prediction error signal e. Thereafter, by preparing two sets each of a predictor, an exclusive OR circuit, and a prediction matching rate counting circuit at the same time as the encoding device, two types of prediction error signals (for the three pixels preceding the decoded pixel) are generated. Create. Furthermore, at the same time as the encoding device, a prediction matching rate comparison circuit determines which of the two types of prediction signals (of the decoded pixel) was used in the encoding device, and a predicted value is selected based on that. The coded image signal l ₀ is obtained by exclusive ORing the decoded coded prediction error signal e.

In the above embodiment, the input image signal a is further binarized, but a multi-tone signal (or analog signal) is used, and two types of binary signal switch A binarization circuit may be provided immediately before one of the inputs (for a binary image).

In addition, in the above embodiment, a case was explained for a so-called predictive coding method, but a prediction error signal sequence is divided into a plurality of sequences based on reference pixel or dither threshold information, and a code suitable for the statistical properties of each sequence is used. The present invention can also be applied to an encoding method based on the so-called Marco model.

〔Effect of the invention〕

As explained above, the present invention uses a low-frequency image signal encoding device that performs average value processing of a predetermined area near a pixel on an image signal containing a mixture of binary images and halftone images. The binary image and the halftone image are separated using a pass-through filter and a comparison circuit that detects and compares the difference between the maximum and minimum values of the output of the low-pass filter within the area, and The output of the low-pass filter is dithered by a dithering circuit, and using each predictor for binary images and dithered images, the prediction matching rate of the preceding predetermined number of encoded pixels is calculated as follows. Since each of the predictors is selectively used to predictively transform and encode encoded pixels, encoding adapted to each pixel can be easily performed, and data can be compressed with high overall efficiency. In addition, it has the excellent effect of being able to reproduce moiré-free, high-quality images with extremely good gradation reproducibility even for halftone photographs using various screens.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing an example of reference pixel positions used for predictive encoding in a conventional image signal encoding device, and FIG. 2 is an example of a prediction method in the image signal encoding device of FIG. 1. 3 is a diagram showing a code example when encoding a prediction error signal in the image signal encoding device of FIG. 1, and FIG. 4 is a diagram showing a conventional image signal encoding device. FIG. 5 is a diagram showing each code example for explaining the operation of the image signal encoding device shown in FIG. 4, and FIG. 7 is a block diagram showing the encoding device; FIG. 7 is a diagram showing pixel positions where average value processing is performed in the image signal encoding device of FIG. 6; and FIG. 8 is a diagram showing the code of the image signal of FIG. 6. FIG. 9 is a diagram showing a dither threshold matrix in the image signal encoding device of FIG. 6, and FIGS. FIG. 7 is a diagram showing each code example for explaining the operation of the image signal coding device shown in FIG. 6; In the figure, 1...predictor, 1a...predictor for binary images, 1b...predictor for dithered images, 2...
...Exclusive OR circuit, 2a...Exclusive OR circuit for binary images, 2d...Exclusive OR circuit for dithered images, 3...Encoder, 4...Low pass filter, 5... ...Maximum/minimum difference counting/comparison circuit within area, 6
...Dither circuit, 7...Binary signal switch,
8a...Binary image prediction matching rate counting circuit, 8b...
. . . Prediction matching rate counting circuit for dithered images, 9 . . . Predicting matching rate comparison circuit, 10 . . . Prediction error signal selector. In each figure, the same reference numerals indicate the same or equivalent parts.

Claims (1)

[Claims] 1. For an image signal in which a binary image consisting of characters, line drawings, etc. and a halftone image are mixed, the pixel signal is predicted based on the values of preceding adjacent pixels, and the prediction error signal is encoded. The predictive coding device includes a low-pass filter that performs average value processing in a predetermined area near the pixel, and the output signal from the low-pass filter is binarized by dithering processing using a variable threshold. detecting the maximum and minimum values of the outputs of the low-pass filter corresponding to pixels in a predetermined area in the vicinity of the pixel, and calculating the difference between them as a predetermined value. a comparison circuit that determines whether or not it exceeds the limit; a binary signal switch that selects either the binary signal or the output of the dithering circuit based on the output signal of the comparison circuit; and the output of the binary signal switch. first and second predictors that receive signals as input and perform predictions for the binary image and the dithered image signal, respectively; and output signals from the first and second predictors, respectively. and first and second predicted matching rate counting circuits that count predicted matching rates in a predetermined number of preceding pixels, and output signals of the first and second predicted matching rate counting circuits. and a prediction error signal selector for selecting one of the two types of prediction error signals.