JPH0626439B2 - Image quality improvement device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to an image input device used for facsimiles, image scanners, and the like, and particularly to a high image quality suitable for an image input device that requires high image quality. Regarding the chemical conversion device.

[Conventional technology]

As a conventional device of this type, for example, there is one disclosed in Japanese Patent Laid-Open No. 58-175364. In such a configuration, an image read from a scanner has a sharpness (MTF) due to the characteristics of an optical sensor. ) Deterioration occurs. Therefore, when fine characters and thin lines are binarized, blurring and blurring occur and the image quality deteriorates.

A method for preventing this deterioration has been proposed for a long time, and a method using a differential filter is generally used. By using this differential filter, it is possible to greatly reduce the deterioration. However, if the above-described correction is performed on an input image such as a photograph that has a continuous gradation change, the gradation change is emphasized and the smoothness and softness of the image are lost. For this reason, when there is no significant change in the pixel level in a certain pixel set, the above-mentioned correction is not performed.

In the conventional apparatus shown in FIG. 7, the detection unit 41 determines whether or not there is a significant difference in gradation change of the signal input from the scanner 1, and the output calculated by the differentiation unit 2 is selected according to the output of the detection unit 41. , The output which is not processed is selected or is switched by the select 5. After that, the binarization unit 6 binarizes the systematic dither method, a constant threshold value, and the like, and the encoding unit 7
Was encoded and transmitted.

[Problems to be solved by the invention]

However, in this conventional apparatus, no consideration was given to the deterioration of the image quality of the halftone image. For this reason, there is a problem that a differential operation is applied to the halftone dot portion, and the image quality is greatly deteriorated.

As shown in FIG. 8, when a character or a photograph is input, fine processing such as a character can be reproduced well by applying a differentiation process. Since the change in the points is emphasized, the image quality is deteriorated.

Therefore, by providing a differential filter and a smoothing filter and selectively using these, it is possible not only to correct the image quality deterioration for fine characters and thin lines, but also for the halftone image. However, this alone causes a problem that the resolution deteriorates when the read density of the input image changes.

An object of the present invention is to provide an image quality improving apparatus which can always obtain a high quality image even if the reading density of an input image changes.

[Means for solving problems]

The object is to convert optical information on an input document into a digital signal by scanning, a differentiating unit for performing a differential operation on a pixel of interest in the digital signal and pixels around the pixel of interest, and the digital signal. A smoothing unit that performs a smoothing operation on the target pixel inside and pixels around the target pixel; and a detection unit that detects the type of the input image from the digital signal in the window among the digital signals,
In an image quality improving apparatus including a switching unit that selects the output of the smoothing unit and the differentiating unit according to the detection result of the detection unit,
A window size for calculating the rate of change of the signal level in the digital signal, changing the size of the window according to the rate of change, and changing the sizes of the smoothing section and the differentiating section in synchronization with the change of the window size. This is achieved by providing a change unit.

[Action]

When the reading density is increased, the size of the smoothing part that smoothes the input halftone dot image is also increased, so the resolution is high,
High-quality images with no moire can be obtained.

〔Example〕

 Hereinafter, the image quality improving apparatus according to the present invention will be described in detail.

FIG. 1 is a block diagram showing an embodiment of the present invention, which comprises a scanner 1, a differentiation unit 2, a detection unit 4, a binarization unit 6, an encoding unit 7 and a selector 51.

The original read by the scanner 1 is photoelectrically converted into a digital signal. This digital signal is supplied to a differentiating section 2 for performing a differential filter, a smoothing section 3 for performing a smoothing filter, and a detecting section 4. The detection unit 4 detects a halftone dot portion and a character portion of the input image, controls the selector 51 so as to select the output of the smoothing unit 3 in the case of the halftone dot portion, and detects the halftone portion and the character portion in the differentiating unit 2 in the case of the character portion. Select an output, otherwise
The signal from Sukiyana is output as it is.

The selected output signal is binarized by the binarization unit 6 and encoded by the encoding unit 7.

By performing the smoothing in this way, the halftone image can be averaged and the level of the original image before halftoning can be reproduced. By converting this original image level into a dither again, it is possible to create a high-quality image without moire.

2 (a) and 2 (b) show the principles of the differential filter, the smooth filter, and the differential filter. Here, an example in which the coefficient is applied to only the pixel of interest and the upper, lower, left, and right sides of the pixel of interest has been described, but it is also possible to select all or some of the eight surrounding pixels and perform the operation. Further, it is possible to change the filter matrix instead of 3 × 3, such as 4 × 3, 4 × 4, 2 × 2, etc., and change the coefficient.

Here, it is considered that two independent filters, that is, a differential filter and a smooth filter, are required. Actually, in the example of FIG. 2, the hardware can be greatly simplified by a few operations as shown in FIG. Is possible.

That is, a circuit that outputs 0.5 times the peripheral pixels including the target pixel (for example, the value stored in the register is 0.5).
The multiplication operation is a simple operation of shifting one bit to the right. ) Is prepared, a subtraction operation is performed by subtracting the input value of the target pixel from the quadruple value, and a smoothing operation is performed by adding the input value of the target pixel to the doubled value. In order to match the level where processing is not performed in both the differential operation and the smoothing operation, a division circuit is required later, but it is possible to omit it or substitute it by a simple shift operation.

FIG. 4 shows a configuration example of the differential / integral filter.

Input C is a digital signal from Scanana. The sum calculator 11 calculates the sum including peripheral pixels, and the output of the coefficient multiplier 10 and the subtractor 12 and the adder 13 calculate the sum. As described above, the output of the subtractor 12 is the differential output and the output of the adder 13 is the smooth output.

The output of the total calculator 11 is also supplied to the delay circuit 15 and the comparator 14. Total calculator 1 delayed by delay circuit 15
If a significant change is not detected by comparing the output of 1 and the current output and the differential output is large, it is determined to be the halftone dot portion and the integral output is selected. On the other hand, if there is a significant change in the output of the sum calculator 11 and the differential output is large, it is determined to be the character portion and the differential output is selected. No other processing is performed. This selection is made by the select 16.

FIG. 5 is an implementation example of the binarization unit 6. The present invention relates to high image quality, and whether or not to perform binarization is not essential, but it is particularly effective when performing binarization for expressing similar halftones. Therefore, a binarization example will be described. To do.

The systematic dithering method is currently widely used to represent the halftone level in binary. For details of this principle, please refer to Takahiko Fukibuki, "Image processing for FAX and OA."
Please refer to p16-p26 published by Nikkan Kogyo Shimbun. Further, it also has means 72 for generating a constant threshold value for transmitting the binary image, and switching is performed by the selector 73 according to the dither threshold value 71 and the select signal S. The select signal S may be designated by a person from the panel surface, or the characteristics of the image signal may be extracted and automatically switched.

The signal selected as described above and the signal created in FIG. 4 are compared by the comparator 74 to obtain the output U.

By improving the image quality in this way, it was found that the halftone dot image was also greatly improved.

FIG. 6 shows another embodiment using the present invention. This corresponds to the case where the read density is increased to 16 lines / mm or the like. When the reading density is improved, it is necessary to increase the size of the smoothing filter for smoothing the input halftone dot image. That is, if the window size for calculating the average is smaller than the halftone dot period of the input image, the average value fluctuates, and the correct switching operation cannot be performed.

For this reason, the average calculation window that determines the switching operation is made variable, and in synchronization with this, the sizes of the smoothing filter 3 and the differential filter 2 are made variable. This size is determined by changing the rate of change of the input image signal level by the window size changing unit
Calculated according to In this case, a smoother control can be performed by giving a hysteresis to the control or by correcting and predicting the change rate of the peripheral pixels.

〔The invention's effect〕

According to the present invention, it is possible to obtain a high-quality image having high resolution and free of moire regardless of the reading density of the input image.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention, and FIG.
(a), (b) is a processing explanatory view of the filter applied to the present invention,
FIG. 3 is an explanatory view of the principle of the differential filter and the integral filter,
FIG. 4 is a block diagram showing the specific construction of the filter, and FIG.
FIG. 7 is a detailed block diagram showing the details of the binarization unit 6, FIG. 6 is a block diagram showing another embodiment of the present invention, FIG. 7 is a block diagram showing a conventional image input device, and FIG. It is a processing image figure by a device. 1 ... Sukiana, 2 ... differential part, 3 ... smooth part, 4 ...
Detection unit, 6 ... Binarization unit, 7 ... Encoding unit, 51 ... Selector.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Tomoaki Aoki 3-2-1, Saiwaicho, Hitachi City, Ibaraki Hitachi Engineering Co., Ltd. (56) References JP-A-57-24166 (JP, A) Special Features Kaisho 58-75374 (JP, A)

Claims (1)

[Claims] 1. A means for converting optical information on an input document into a digital signal by scanning, a differentiating unit for performing a differential operation on a pixel of interest in the digital signal and pixels around the pixel of interest, and the digital unit. A smoothing unit that performs a smoothing operation on the pixel of interest in the signal and pixels around the pixel of interest; a detector that detects the type of the input image from the digital signal in the window among the digital signals;
In an image quality improving apparatus including a switching unit that selects the output of the smoothing unit and the differentiating unit according to the detection result of the detection unit,
A window size for calculating the rate of change of the signal level in the digital signal, changing the size of the window according to the rate of change, and changing the sizes of the smoothing section and the differentiating section in synchronization with the change of the window size. An image quality improving device characterized by having a changing unit.