JPS60206373A - Binarization system - Google Patents

Abstract

PURPOSE:To realize the constitution of a device which prevents a binarization error, has a good reproducibility of a thin line in the moving direction of a notice picture element, can execute a high speed processing, and also is inexpensive, by determining a binarization threshold by using a weighted average value of a representative value of a density level of a notice picture element group and its adjacent picture element group. CONSTITUTION:Plural picture elements arranged in the sub-scanning direction forms one picture element group. A representative value of a density level of a notice picture element group read out of a line buffer 11 is calculated by a picture element group representative value calculating circuit 12. A weighted average calculating circuit 13 calculates a weighted average value of the notice picture element group by using an output of the calculating circuit 12 and an initial value S, when the head picture element group becomes the notice picture element group, and by using the output of the calculating circuit 12 and a weighted average value to the picture element group which is immediately before being read out of a storing circuit 14, when the subsequent picture element group becomes the notice picture element group. An optimum threshold operating circuit 15 executes an equation I by using an output of the calculating circuit 13, calculates a binarization threshold Th to the present notice picture element group, and inputs it to a binarization circuit 16. In this case, a picture signal of its notice picture element group is synchronized by a delay synchronizing circuit 17, inputted to the binarization circuit 16 and binarized.

Description

[Detailed Description of the Invention] [Technical Field] The present invention provides two
Regarding valorization method.

[Prior art]

When an image read by scanning a document with a pixel resolution scanner is binarized, if the binarization threshold is fixed, it cannot follow the background density level of the document, and binarization errors are likely to occur. Conventionally, the image signal is integrated by an analog integration circuit to obtain a binarized threshold voltage that follows the background level of the original.

However, with such a method, it is impossible to make the binarization threshold faithfully follow dynamic fluctuations in the background density level of the document, and binarization errors cannot be sufficiently prevented.

Further, the response characteristics of the analog integrating circuit are easily affected by changes over time and variations in the elements, and as a result, the binarization operation tends to become unstable.

In order to solve such problems, a method of dynamically determining the binarization threshold value using digital processing can be considered. For example, a method has been considered in which a binarization threshold for a pixel of interest is calculated from the density levels of neighboring pixels on the same line as the pixel of interest.

However, such a method has the problem that the amount of calculation processing increases because calculations are performed on a pixel-by-pixel basis, and high-speed processing is not easy. Furthermore, there is a serious problem in that the reproducibility for thin lines in the movement direction of the pixel of interest, that is, in the main scanning line direction, is reduced.

〔the purpose〕

It is an object of the present invention to prevent binarization errors caused by dynamic fluctuations in the background density level of a document, to have good reproducibility of thin lines in the moving direction of the pixel of interest, and to reduce the amount of calculation processing required to determine the binarization threshold. The object of the present invention is to provide a new binarization method that enables high-speed processing and realizes an inexpensive device configuration.

〔Example〕

Hereinafter, one embodiment of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a conceptual diagram for explaining image division in this embodiment. In this figure, each square grid represents a pixel. In this embodiment, the image is divided into four pixels arranged in the sub-scanning direction as one pixel group. The pixel group of interest moves sequentially in the main scanning direction.

To determine the binarization threshold, first, a weighted average value X is obtained by weighting the representative values of the pixel group of interest and its neighboring pixel groups in accordance with the exponential distance. I put it on.

Specifically, it is calculated using equation (1).

Xo= (1-m)Xo0mx' (],) where,
Xn is the representative value of the pixel group of interest, X' is the weighted average value X determined when the previous pixel group of the current pixel group of interest was the pixel group of interest, and 9m is a weighting parameter.

The weighted average value X of the pixel group of interest thus determined. The binarization threshold Th for each pixel belonging to the pixel group of interest is calculated using the following equation (2).

T, = aXo+β (2) Here, α and β are constants.

In this way, in this embodiment, since the calculation process is performed using four pixels as a unit, the amount of calculation processing is reduced to one fourth or less compared to the case where the unit is -2 pixels.

Note that the above recurrence formula (+
, ) is also effective in reducing the amount of calculation processing.

=3-Here, when the first pixel group becomes the pixel group of interest, an appropriate value is initially set as X' in the calculation of the weighted average value. Furthermore, in the last part of the document, the number of pixels may be insufficient and it may not be possible to divide the document into groups of four pixels. In this case, a pixel group is formed using the existing number of pixels and processed.

Now, the density level of a thin line in the main scanning direction decreases due to deterioration due to the dot spread of the scanner, so if the binarization threshold is determined only based on the pixel density level in the main scanning direction, the problem will arise that the thin line will not be reproduced. easy. However, according to this embodiment, the representative value of the density level of only four pixels including the thin line is lower than the density level of the pixels forming the thin line.

The binarization threshold value calculated based on the representative value is never greater than the density level of the thin line. Therefore, deterioration in reproducibility of thin lines in the main scanning direction can be avoided.

FIG. 2 is an explanatory diagram thereof. In this figure, 1 is a thin line with a width of one pixel in the main scanning direction, and 2 is a waveform indicating the density of the thin line 1 on the document. 3 is a waveform showing the density level of the thin line 1 after being read by the scanner, and it can be seen that the density level decreases due to the point spread of the scanner.

4 is a representative value of the density level of each pixel group (four pixels) including thin line 1, and 5 is a binarization threshold determined based on the representative value.

In this case, the MTF of the scanner is 0.29, and its impulse response (point spread function) is as shown in FIG.

FIG. 4 is a block diagram of the apparatus according to this embodiment. In this figure, numeral 11 is a line buffer that can store image signals for eight lines. The image signal (Kukata pixel) obtained by digitizing the analog image signal obtained by the scanner through an analog/digital converter is serially input to the line buffer 1 for each main scanning line and is sequentially accumulated there. , and the image signals accumulated in parallel thereto are sequentially read out pixel by pixel.

12 is a pixel group representative value calculation circuit that calculates the representative value of the density level of the pixel group of interest read out from the line buffer 11, and the requested representative value is given to one input of the average calculation circuit 13 with a weight of -6. . When the first pixel group becomes the pixel group of interest, the weighted average calculation circuit 13 calculates a weighted average value for the pixel group of interest using the representative value input from the pixel group representative value calculation circuit 12 and the initial value S. ,Output. This weighted average value is stored in the storage circuit 14. When the pixel group after that becomes the pixel group of interest,
The weighted average value of the pixel group of interest is calculated by inputting the representative value input from the pixel group representative value calculation circuit 12 and the weighted average value for the pixel group just before being read out from the storage circuit 14.

15 is an optimal threshold calculation circuit, and a weighted average value calculation circuit 1
The above formula (2) is executed using the weighted average value input from 3, and the binarization threshold for the current pixel group of interest is calculated.
It is input to the value conversion circuit 16. At this time, the image signals of the pixel group of interest are synchronized by the delay synchronization circuit 17 and input to the binarization circuit 16.

The binarization circuit 16 operates for each pixel in the pixel group of interest.

The image signal is binarized using a binarization threshold. The signal binarized at the middle of the pixel group by the binarization circuit 16 is temporarily stored in the line buffer 18 and outputted to the outside as a serial signal for each main scanning line.

Note that in the above description, the pure average value is used as the representative value of the pixel group, but the sum of the density levels of each pixel of the pixel group may be used. In this case, the calculated binarization threshold is set to 4
Divide by. Further, although four pixels are used as one pixel group, the number of pixels in the pixel group is not limited to this.

〔effect〕

As described above, according to the present invention, the weighted average value of the representative values of the density levels of the pixel group of interest and its neighboring pixel groups is used to calculate the binary value for each pixel in the pixel group of interest using a specific arithmetic expression. By determining the threshold value, the following effects can be achieved.

(1) Dynamic fluctuations in the background density level of the original 2
This is because the valuation threshold can be tracked.

Binarization errors that occur in systems that use analog integration circuits to determine the binarization threshold are less likely to occur.

(2) The calculation for determining the binarization threshold value can be performed for each pixel group, and the amount of processing is small, so high-speed processing is possible and can be easily performed using a simple and inexpensive device.

(3) Sufficient reproducibility can be obtained for thin lines in the moving direction of the pixel of interest.

[Brief explanation of the drawing]

Figure 1 is a conceptual diagram to explain the pixel group division of an image, Figure 2 is an explanatory diagram regarding the reproducibility of thin lines parallel to the moving direction of the pixel of interest, Figure 3 is an illustration of impulse response, and Figure 4 FIG. 1 is a block diagram showing a binarization device according to an embodiment of the present invention. 11.18... Line buffer, 12... Pixel group representative value calculation circuit, 13... Weighted average calculation circuit, 14...
・Storage circuit, 15... Optimal threshold calculation circuit, 16...
Binarization circuit, 17...delay synchronization circuit. 8- Input Fang 1 figure Fang 2 figure 4 figure,! , signal line l ma", ke2 5 Okuchi 1ko grudge 14 [D Shito1 !-, υ mouth lj 5; 埒Pay 17 801 row a 15 # taku1koropo<>S Rimap

Claims (1)

[Claims] (1) In the binarization method of an image read by a pixel decomposition type scanner, the image is decomposed into consecutive pixel groups for each of multiple pixels across multiple lines, and the density level of the pixel belonging to each pixel group is used. The representative value of the density level of the pixel group is determined by performing a specific calculation using A binarization method characterized in that a binarization threshold value for a pixel belonging to a pixel group of interest is determined by performing the calculation.