JP3475606B2 - Image processing device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing device in an image reading device such as an image scanner or a copying machine.

[0002]

2. Description of the Related Art In the conventional processing method of this type, as shown in FIG.
Bit (that is, the gradation of each pixel is represented by 8 bits)
The read data is subjected to preprocessing such as shading correction and then passed to the color correction / enlargement / reduction unit 1. The color correction / enlargement / reduction unit 1 outputs 8-bit data after performing color correction and image size enlargement / reduction. This 8-bit data is passed to the area determination unit 2, the character binarization unit 3, and the γ correction unit 4, respectively.

The area determining unit 2 selects the 8-bit data from among the 8-bit data.
For example, using the upper 4-bit data, it is determined whether the image area is a character area or a photograph area. Here, the character area means an area composed of only two tones of white for the background and black for the character. Further, the photographic area is an area represented by a large number of gradations from white to black.

The character binarizing unit 3 is for converting a character area into 1-bit data, and uses a certain specific data such as the most significant 1-bit of the input 8-bit data. , This binarization processing is performed.

The γ correction unit 4 is for performing γ correction on the photographic area, and performs this γ processing on 8-bit data. The data output from the γ correction unit 4 is then sent to the halftoning unit 5, where it is converted into 1-bit data by a halftoning method such as a dither method, a minimum average error method, or an error diffusion method.

Based on the signal from the area determining section 2, the selector 6 outputs the signal from the character binarizing section 3 if the area is a character area, or the signal from the halftoning section 5 if the area is a photographic area. Is selected and binarized and output.

[0007]

[Problems to be Solved by the Invention]
One problem is that the accuracy of 8-bit data input by the γ correction unit 4 is substantially reduced to 6-bit data.

That is, the 8-bit data input to the γ correction unit 4 is data representing 256 gradations from 0 to 255, which is converted according to the nonlinear γ curve by γ correction, and The resolution for the key is substantially reduced to 6 bits, that is, about 64 gradations.

That is, the output of the γ correction unit 4 is actually 8 bits, but the amount of information held by the data is substantially the amount of information of about 6 bits.

As described above, the gamma correction unit 4 lowers the resolution for gradation by about 2 bits, and as a result, there is a problem in that the image quality of the photographic area in the output data deteriorates.

With this conventional configuration, if it is desired to give substantially 8-bit data to the halftoning section 5,
It is necessary to increase the original data input to the color correction / enlargement / reduction unit 1 to 10 bits.

However, in doing so, the color correction / enlargement / reduction unit 1
Will also have to be 10 bits. This color correction
Since the scaling process is a relatively complicated process,
If the configuration for 10 bits is changed to the configuration for 10 bits, there is a problem that the cost increases.

Therefore, an object of the present invention is to provide an image processing apparatus capable of performing image data processing with an improved image quality of a photograph portion without increasing the cost by using a color correction / enlargement / reduction unit having the same structure as the conventional one. To provide.

[0014]

The image processing apparatus of the present invention first performs γ correction on original image data, and then performs color correction and enlargement / reduction processing. The image data that has been subjected to the color correction / enlargement / reduction processing is provided on the one hand so as to be used for, for example, the halftoning processing of the photographic region, and on the other hand, 1 / γ which is the inverse conversion of the γ correction. The image data is corrected and provided for processing such as binarization of a character area and area identification of a character area and a photo area.

According to the image processing apparatus of the present invention, the first γ
The accuracy of the image data is reduced by about 2 bits at the correction stage,
Color correction / enlargement / reduction processing is performed on the data with the lowered accuracy. Furthermore, 1 / γ correction is performed on the data that has been subjected to color correction / enlargement / reduction processing, and the accuracy of the data is further reduced by about 2 bits. The halftoning process of the photo area can be performed on the data after the color correction / enlargement / reduction process, that is, the data whose accuracy is reduced by about 2 bits from the original data. On the other hand, the binarization processing of the character area and the discrimination processing of the character area and the photograph area can be performed on the data after the 1 / γ correction, that is, the data whose accuracy is reduced by about 4 bits from the original data.

Therefore, if the data to be subjected to the half-toning processing needs to have an accuracy of 8 bits, for example, the original image data must be at least 10 bits.
In that case, since the color correction / enlargement / reduction processing only needs to be performed on the 8-bit data after the γ correction, the cost increase unlike the conventional case does not occur. Further, the character region binarization process and the region identification process are performed on the data with lower precision, but since these processes do not originally require high precision data, there is a problem. Can be processed without.

[0017]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings.

FIG. 2 is a block diagram of an image processing apparatus showing an embodiment of the present invention.

In the present invention, the color correction / enlargement / reduction unit 1
In order to keep the number of bits of the image data input to the halftoning unit 5 to be 8 bits, the γ correction unit 4 is provided in the color correction / scaling unit 1 as shown in FIG. It is arranged in the preceding stage, and 10-bit image data is input to the γ correction unit 4.

Further, after the color correction / enlargement / reduction unit 1, before the region determination unit 2 and the character binarization unit 3, a 1 / γ correction unit 7 for performing inverse conversion of γ correction is newly provided. It is a thing. Other configurations are the same as those of the conventional example shown in FIG.

In such a configuration, 10-bit data read by an image sensor (not shown) and subjected to preprocessing such as shading correction is first passed to the γ correction unit 4. The γ correction unit 4 converts this 10-bit data according to a predetermined γ curve. As described above, this γ correction causes a decrease in accuracy of about 2 bits. Therefore, the γ correction unit 4 expects the accuracy decrease of about 2 bits and drops the input 10-bit data to 8-bit data for output. The 8-bit data is then sent to the color correction / enlargement / reduction unit 1 and subjected to color correction processing and enlargement / reduction processing as in the conventional case.

The 8-bit data output from the color correction / enlargement / reduction unit 1 is used as a 1 / γ correction unit 7 and a halftoning unit 5.
Passed to. The 1 / γ correction unit 7 performs the inverse conversion of the γ correction and converts it into data having the same characteristics as the original 10-bit data. At this time, since non-linear conversion is also performed,
The accuracy of the input 8-bit data is further reduced by about 2 bits.

However, the 1 / γ correction unit 7 further reduces the precision-decreased data to 4 bits and outputs it. The reason is that the 4-bit output data is sent to the area determination unit 2 and the character binarization unit 3 next, but the area determination unit 2 and the character binarization unit 3 originally output data with high gradation resolution. The area determination unit 2 does not need the 4-bit, character binarization unit 3
Is because 1-bit data is sufficient.

That is, the area determination unit 2 and the character binarization unit 3
Since the input data does not originally require such a high gradation resolution, even if the gradation resolution of the data, that is, the accuracy is lowered by the γ correction and the 1 / γ correction, it does not cause any problem.

The area determination unit 2 determines whether the area is a photograph area or a character area based on the input 4-bit data. Character 2
The binarization unit 3 outputs 1-bit character data by using the most significant 1 bit of the input 4-bit data.

The 8-bit data output from the color correction / enlargement / reduction unit 1 is also sent to the halftoning unit 5. The half-toning unit 5 outputs 1-bit data represented by gradation using a half-toning method such as a dither method, a minimum average error method, or an error diffusion method based on the input 8-bit data. In the halftoning unit 5, since the input data is 8-bit data, that is, data representing 256 gradations, it is possible to output photographic data with high image quality.

The selector 6 selects and outputs the character data from the character binarization unit 3 or the photographic data from the halftoning unit 5 based on the signal from the area determination unit 2.

According to this embodiment of the present invention, since the gradation is expressed using 8-bit data in the photographic area, the image quality is not deteriorated. Compared with the conventional configuration, 1
/ Γ correction unit 7 is increasing, but this is 256 words × 4
Since it can be realized by a ROM with a small bit, the cost does not increase so much.

Further, in order to obtain the same image quality in the conventional system, the color correction / enlargement / reduction unit 1 had to be 10 bits, but in the embodiment of the present invention, the color correction / enlargement / reduction is performed. The section 1 can be applied with the 8-bit configuration as it is.

Therefore, there is no increase in cost due to the enlargement of the color correction / enlargement / reduction unit 1. In addition, the γ correction unit 4
Is converted to 10 bits, but this is not required to be a large increase in cost because it is only necessary to configure the ROM.

As a result of the above, high image quality of photographs can be obtained without increasing the cost.

[Brief description of drawings]

FIG. 1 is a block diagram of an image processing apparatus according to a conventional example.

FIG. 2 is a block diagram of an image processing apparatus according to an embodiment of the present invention.

[Explanation of symbols]

1-color correction / scaling unit 2 area determination section 3 character binarization part 4 γ correction unit 5 Half-toning section 6 selector 7 1 / γ correction unit

Claims (2)

(57) [Claims] 1. A and γ correcting unit that performs correction γ on the image data, and color correction and scaling unit for performing color correction and enlargement / reduction processing on the image data outputted from the γ correcting unit, before
The image data output from the color correction / enlargement / reduction unit, and a 1 / γ correction unit that performs inverse conversion of γ correction to the image data output from the color correction / enlargement / reduction unit; image to provide the image data output from / gamma correction section
In the image processing device, a character binarization unit that binarizes a character area of the image data output from the 1 / γ correction unit, and a photographic area of the image data output from the color correction / enlargement / reduction unit to 2 A half-toning unit for digitizing, a region determining unit for identifying a photograph region and a character region of the image data output from the 1 / γ correcting unit, and the character binarizing based on a signal from the region determining unit. the image processing device for a selector for selecting the output image data from one of the halftoning section and parts, comprising the. 2. The image processing apparatus according to claim 1,
The γ correction unit displays the number of gradations required for the photographic area.
Image data with more bits than necessary
It can be input and converted into image data with the required number of bits.
The image processing apparatus according to claim and.