JPH08116444A - Binarized image processing unit - Google Patents

Abstract

PURPOSE: To obtain the binarized image processing unit in which reduction in a texture is realized without addition of a special circuit. CONSTITUTION: The binarized image processing unit is made up of an image sensor 1 to receive an image, an A/D converter 2 quantizing an image signal outputted from the sensor 1 to an n-bit digital signal, an error diffusion processing section 3 binarizing the digital signal outputted from the A/D converter 2 and a block 4 dithering a white reference level WH signal and a black reference level BL signal within a prescribed amplitude from a certain value and providing outputs of them synchronously with a received image.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a binarized image processing apparatus used in facsimiles, copiers and the like, and more particularly to a binarized image processing apparatus using an error diffusion method.

[0002]

2. Description of the Related Art Conventionally, when a continuous image signal is converted into a discrete image (hereinafter referred to as an original image) signal and then converted into a binary digital image (hereinafter referred to as a binary image) signal, error diffusion is performed. There is a method of obtaining a pseudo gradation by the method. This method is described by Floid and Steinberg in 1976 as "An Adaptive Algorithm for Spatial Grayscale".
(Proceeding of the SID Vol. 17/2 Second
This is proposed by a paper called Quater 1976) and is characterized by preserving the image density by dispersing the error with respect to the original image generated by the binarization process to surrounding pixels.

Further, as a method for varying the brightness of an output image in a binarized image processing apparatus using this error diffusion method, the present inventors have previously filed Japanese Patent Application No. 6-106221.
No. 3, proposes a method in which the brightness of the output image can be easily changed by changing the white reference level and the black reference level used when calculating the error.

By the way, in the binarized image processing apparatus using the error diffusion method, when an image of medium brightness and a little level change is processed, a unique striped pattern (hereinafter referred to as texture) is generated in the binary image. There is a problem. This texture is due to the fact that the binary signal changes periodically at high frequencies. Therefore, in order to reduce the texture, it is necessary to process the multi-valued data including the current image data and the error data so that the binary signal does not change periodically.

As a technique for reducing such texture, the technique disclosed in, for example, Japanese Patent Laid-Open No. 3-34776 has been conventionally used. The method disclosed in this publication reduces the texture by disturbing the periodicity of data by changing the threshold value used for binarization. This method uses the above-mentioned error diffusion method of variable output image brightness. When combined with a binarized image processing device, the configuration is as shown in the block diagram of FIG. In FIG. 3A, 101 is an image sensor such as a CCD that functions as an image capturing means, and 102 is an A that converts a signal obtained from the image sensor into an n-bit digital signal.
/ D converter, 103 is an error diffusion processing unit that converts an n-bit digital signal into a binary digital signal, and 104 is a value WH (white reference level) for varying the output image brightness, BL
A block for outputting (black reference level) and a threshold generation block 105. Then, in the error diffusion processing unit 103, the variable threshold output from the threshold generation block 105 is used to reduce the texture.

As another technique for reducing the texture, Japanese Patent Application No. 3-88570 discloses a technique for reducing the texture by changing the weighting coefficient for distributing the error. When this method is combined with a binarized image processing apparatus using an error diffusion method with variable output image brightness, the configuration shown in the block diagram of FIG. 3B is obtained. In FIG. 3B, 101 to 104 are image sensors, A / D converters, error diffusion processing units, and WH · B.
The L output block has the same configuration as the binarized image processing apparatus shown in FIG. 3A, and 106 is a weighting coefficient variable block. Then, in the error diffusion processing unit 103, the texture is reduced by using the variable weight coefficient output from the weight coefficient variable block 106.

In addition to this, as means for reducing texture, a method of changing the error matrix size, a method of adding a numerical value to the original image data itself, a method of switching the processing method to the dither method, a dot of an output printer Various methods are known, such as a method of changing the size.

[0008]

However, the conventional binarized image processing apparatus configured to perform texture reduction with the above-described structure must add a separate circuit or the like only for texture reduction, and the circuit It is necessary to increase the scale and the number of parts.

The present invention has been made in order to solve the above-mentioned problems in the conventional binarized image processing apparatus capable of reducing texture, and reduces texture without adding special circuits and parts. It is an object of the present invention to provide a binarized image processing device capable of realizing the above.

[0010]

In order to solve the above problems, the present invention relates to an image processing apparatus for binarizing a digitized original image by an error diffusion means, in which pixels of a line before a target pixel are processed. Memory means for holding error data generated when the data is binarized, means for weighting the error data held in the memory means, data from the error data weighting means, and the notice A unit for adding the pixel data to obtain the corrected image data, a unit for setting the first reference level and the second reference level for making the brightness of the output image variable, and the corrected image data with a predetermined threshold value. And a means for calculating error data using the corrected image data and the first and second reference levels, wherein the first and second reference level setting means are The second reference level, and outputs to vibrate in synchronization with each input image signal, and configured to reduce the texture.

[0011]

[Operation] The error data is changed by vibrating the first reference level and the second reference level for varying the brightness of the output image in synchronization with the input image signal. Since the periodicity of the value signal can be disturbed, the texture can be reduced without adding a separate circuit or the like.

Next, this point will be specifically described.
When the n-bit digital signal input from the A / D converter that digitizes the subject image is subjected to error diffusion processing, the white reference level WH and the black reference level BL used for calculating the error data are externally controlled. The error data obtained by the calculation is the difference between a signal obtained by adding the error data generated in the peripheral pixels to the image signal (hereinafter referred to as the corrected image signal) and the binary level. For example, when the binarized result is white and WH is the white reference level R (R: image dynamic range) in the conventional error diffusion method, the generated error data E is E = corrected image signal −R. Since E is negative, if it is later added to another image signal, the corrected image signal becomes small. That is, the error data E at this time works so as to blacken the peripheral pixels. Now, when WH is made smaller than R, the resulting error data becomes smaller than E. That is, the degree to which the peripheral pixels are blackened is small. Similarly B
When L is made larger than the black reference level 0 in the conventional error diffusion method, the degree of whitening the peripheral pixels becomes small.

As described above, the white reference level WH (first reference level) and the black reference level BL (second reference level)
By making variable, it is possible to change the brightness of the output image. Therefore, the white reference level WH and the black reference level BL
By vibrating and finely varying, the periodicity of the output image can be disturbed, and as a result, the texture can be reduced.

[0014]

EXAMPLES Next, examples will be described. FIG. 1 is a schematic block configuration diagram showing an embodiment of a binarized image processing device according to the present invention. In FIG. 1, 1 is an image sensor that captures an image, and 2 is an A for quantizing an image signal output from the image sensor 1 into an n-bit digital signal.
An / D converter, 3 is an error diffusion processing unit that binarizes the digital signal output from the A / D converter 2, and 4 is a white reference level WH signal and a black reference level BL signal, which have a certain amplitude from a certain value. It is a block that vibrates inside and outputs in synchronization with the input image.

In the binarized image processing apparatus configured as described above, the subject image is captured by the image sensor 1, converted into an n-bit digital signal by the A / D converter 2, and the error diffusion processing section In 3, the binary data is output. Here, the white reference level WH and the black reference level BL signal are used by the error diffusion processing unit 3 when calculating an error with the original image.

Next, regarding the error diffusion processing unit 3 and the block 4 for vibrating and outputting the WH signal and the BL signal,
This will be described in detail with reference to FIG. In FIG. 2, 11a is a subtractor for subtracting the signal b (WH signal) from the signal a,
Reference numeral 11b is a subtractor for subtracting the signal c (BL signal) from the signal a. 12 compares the signal a and the signal d (threshold value), and a> d
It is a comparator that outputs "L" at the time of, and outputs "H" at other times. 13 is a selector, which is the output signal g of the comparator 12.
Is g = “L”, the output signal e of the subtractor 11a is changed to g
When "H", the output signal f of the subtractor 11b is output as the error signal h. Reference numeral 14 is an error memory for storing an error amount from the white or black reference level calculated by the subtractor 11a or 11b, 15 is a block for weighting the above error, 16 is data of the error memory 14 weighted with the original image signal Is an adder that adds These elements constitute an error diffusion processing unit, and 17 is an output block of the white reference level WH and the black reference level BL (hereinafter abbreviated as a counter block) using a counter, which is shown in FIG. An example of the block 4 for vibrating and outputting the WH signal and the BL signal is shown.

Next, the operations of the error diffusion processing section and the counter block thus configured will be described. First, the operation of the counter block 17 will be described. This block 17 includes a white reference level WH and a black reference level B.
From the value k that sets L to change the output image brightness, ± x
In this range, the output image is periodically vibrated in synchronization with the input image. For example, in order to increase the output image brightness of the 8-bit original image, the white reference level WH is set and k
= 200 and x = 2, white reference level WH
Oscillates cyclically as 198, 199, 200, 201, 202, 198, ... For each input pixel. Here, x is a parameter that affects the output image brightness, so it is desirable not to take a large value.

Next, the white reference level WH and the black reference level B set in the counter block 17 as described above.
The operation of the error diffusion processing unit using L will be described. A
The digital signal output from the / D converter 2 is added by the adder 16 to the signal weighted by the weighting block 15 to which the error data calculated in the past output from the error memory 14 is added, and becomes a signal a. That is, the signal a becomes a corrected image signal for the original image signal, and the signal a is binarized. The subtractors 11a and 11b perform subtraction between the signal a and the oscillating WH and BL signals, and ab and ac, and one of these two results becomes error data generated during binarization. . In the comparator 12, the signal a has a preset threshold value d, for example, 1 / the dynamic range R of the image.
2 is compared, and the result is output as binary data. In the selector 13, the binary data g is white (g =
Depending on whether it is “L”) or black (g = “H”), the error memory 14
It operates to switch the error data output to. That is, when the binarized result is black, the difference data (a-c) between the corrected image signal and the BL signal is output as the error data.

As described above, in the binarized image processing apparatus configured as described above, the white reference level WH
The black reference level BL periodically oscillates for each pixel, and when the white reference level WH and the black reference level BL are changed, the error data changes. As a result, the binarized signal of the output image is also changed. Change. Therefore, when the white reference level WH and the black reference level BL are vibrated with an amplitude that does not affect the brightness of the output image, the periodicity of the output image is disturbed, and the texture is reduced.

Here, as a method of vibrating the white reference level WH and the black reference level BL, for example, in addition to the method using the counter shown in this embodiment, a matrix or the like is used to periodically vibrate. It may be configured, or may be configured to vibrate randomly using a random number or the like.

The white reference level WH and the black reference level BL may be vibrated to be processed or not controlled by an external signal such as the brightness of the input image or image edge information. For example, the processing may be performed only when the input image has medium brightness, or may be configured to perform the processing only when the portion to be binarized is other than the edge.

[0022]

As described above on the basis of the embodiments,
According to the present invention, the first and second reference levels for changing the brightness of the output image are configured to vibrate in synchronization with the input image signal, so that a special additional circuit is not provided, It is possible to provide a binarized image processing device that can reduce texture and is suitable for downsizing.

[Brief description of drawings]

FIG. 1 is a schematic block configuration diagram showing an embodiment of a binarized image processing device according to the present invention.

FIG. 2 is a block configuration diagram showing a configuration example of an error diffusion processing unit in the embodiment shown in FIG.

FIG. 3 is a schematic block configuration diagram showing a configuration example of a conventional binarized image processing apparatus for reducing texture.

[Explanation of symbols]

 1 Image Sensor 2 A / D Converter 3 Error Diffusion Processing Unit 4 WH and BL Oscillating Block 11a, 11b Subtractor 12 Comparator 13 Selector 14 Error Memory 15 Weighting Block 16 Adder 17 WH and BL Output Using Counter block

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Office reference number FI Technical display location H04N 1/40 H04N 1/40 101 E 103 Z

Claims (1)

[Claims] 1. An image processing device for binarizing a digitized original image by an error diffusion means, holding error data generated when binarizing data of pixels on a line preceding a pixel of interest. Memory means for storing, means for weighting the error data held in the memory means, means for adding the data from the error data weighting means and the pixel data of interest to obtain corrected image data, and an output image Means for setting a first reference level and a second reference level for varying the brightness of the image, a means for binarizing the corrected image data with a predetermined threshold value, the corrected image data and the first and second And a means for calculating error data using two reference levels, wherein the first and second reference level setting means shake the first and second reference levels respectively in synchronization with the input image signal. A binarized image processing device, which is configured to be moved and output to reduce a texture.