JPS61123274A - Picture processing system - Google Patents

Abstract

PURPOSE:To obtain a recording picture with high quality by changing the weight at the center value of a Laplacian filter in response to digital picture information to correct the degradation in frequency characteristic caused due to the transmission characteristic of luminous amount versus picture density in a printer. CONSTITUTION:The median Z of the 3X3-constitution Laplacian filter (figure 1 (a)) is positive and a function of digital picture information at a noted picture element D(2,2) being a processing object among a prescribed picture element area extracted in response to the Laplacial filter constitution as shown in a figure (b). Thus, new digital picture information processed by the spatial filter is obtained by multiplying matrix elements corresponding to the Laplacian filter and an input picture element, summing the products and normalizing the sum. Thus, the deterioration in frequency characteristic due to nonlinear transmission characteristics in the luminous amount versus picture density of the printer is compensated and a recording picture with high quality is obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION Technical Field The present invention relates to an image processing process that processes the image information read by a scanner to optimize it for image reproduction. The present invention relates to an image processing method that performs processing according to the characteristics of an electrophotographic printer such as an electrophotographic printer.

BACKGROUND OF THE INVENTION Recently, a digital copying machine has been developed in which a laser printer records and reproduces an image based on image information ζ of an original document read pixel by pixel by a scanner.

FIG. 7 shows the configuration of the laser printer.

In the figure, l is a laser oscillator, 2 is a polygon mirror, 3 is an fθ lens, 4 is a reflection mirror, 5 is a photosensitive drum, 6 is a charger, 7 is a developer, 8 is a transfer/separation charger, 9. It is a cleaner.

In the nt laser printer bottle ζ configured in this way, the photoreceptor surface is exposed to laser output light 6ζ modulated according to the binary image information BS sent from the scanner III,
Formed by that exposure? Image formation is carried out in a so-called electrophotographic process in which the W image is visualized by toner development and transferred to recording paper lO.

However, in such an electrophotographic (C) printer, as shown in FIG. This results in problems such as blurring and skipping of dots, and especially irregularities in gradation and poor midtone reproducibility.Especially in high density areas, image distortion and dots appear in low fIk areas. jumps occur, and high-frequency components of the image are lost.

Conventionally, when a printer performs recording/reproduction based on r/L-order image information read by a scanner, various image processing methods such as shading correction and MTF correction are performed to improve image quality. All of the processing means are only meant to correct image deterioration that occurs when the scanner is reading the image, and even if this erroneous image processing is performed, the amount of light in the printer - the image! A recorded image of high quality cannot be obtained unless the influence of non-linearity of the one-time transfer characteristic is corrected.

Purpose The present invention has been made in consideration of the above points, and is an object of the present invention, which is designed to improve the ratio of light intensity to image density in an electrophotographic printer when an image is recorded and reproduced based on original image information read by a scanner. This invention provides nine image processing systems that can correct the deterioration of frequency characteristics caused by non-linearity of transfer characteristics and obtain high quality recorded images.

Structure In order to achieve the purpose of the present invention, the scanner Cζ is selected according to the non-linearity of the light intensity-image density transfer characteristic of the electrophotographic printer. The frequency characteristics are improved by appropriately changing the weight of the center value of a Laplacian filter that spatially filters digital image information that is read in units of seven pixels and subjected to multilevel quantization.

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

The amount of light on the printer side shown in Fig. 8 6ζ - Image I1
Due to the transfer characteristic of 1 degree, when the amount of light shown in the figure is small i, that is, when the concentration is high, the width of the degree fluctuation is small with respect to the change in the amount of light. There is. Similarly, the characteristics in the portion C where the amount of light is high and Cf1kK is low are also the same. People like this, C
Components with high line image frequencies are compressed in this region, and have a one-sided t transfer characteristic in the low frequency direction. This means that on the recorded image, collapse, skipping, etc. occur, resulting in poor resolution of details and deterioration of gradation. Further, the portion having almost linearity indicated by B in the figure is a region where the image frequency is relatively well transmitted. This process is shown in Figure 9, which shows the transfer characteristic of the image frequency.

These characteristics correspond to areas A, B, and C in the figure, respectively, and areas A, B, and C in FIG.

This method requires an erroneous process to improve the frequency characteristics according to the value of digital image information regarding the fIL degree of the n-th original document image read in units of m elements by the scanner and multivalued quantized.

Therefore, in the present invention, the weight of the center value in the Laplacian filter is changed to the digital image information of the pixel to be processed, and the spatial filter processing with different frequency characteristics is performed by spores. .

FIG. 1(a) shows an example of a Laplacian filter with a 3×3 configuration, where the center value zH plus the value and the same r! i! As shown in J(1)), the target pixel DC2,2)4 is a function of the digital image information, which is the target pixel to be processed from among the fixed pixel area extracted according to the Laplacian filter configuration fζ. The numerical array around the central value 2 remains unchanged. At this time, the surrounding values are also treated as a function of the digital image information of the pixel of interest D (2, 2). The same result can be obtained relative to C
ζ, so only the center value here? Rather than change.

Note that in FIG. 1 (11), X indicates the main scanning direction t by the scanner.

The spatial filter processing performed here is digital calculation using so-called convolution integration, and the calculation formula is as follows.

-2XIXl, 2) + ZXD (2, 2) - 2XD (3,
2)-1XD(1,3J-2XD(2,3)-1XD(
March...(1) That is, by multiplying the corresponding matrix of the Laplacian filter and the input pixel by +7x elements, summing them, dividing by (2-12), and normalizing. Spatial filter processing is performed on the target pixel 6ζ. A new digital image information lit DI (2, 2
) will be obtained.

In this case, the present invention particularly uses jiP for the Laplacian filter.
The value of the weight 2 of the center value of #'j is the erase (2, 2
Jmat is the surrounding 111i ([D(1,1)~D(
3, 3) are appropriately changed according to the averaged value to make the spatial filter processing sequential. Figure 2 shows an example of how the weight 2 of the center value of the Laplacian filter changes with respect to the value of the pixel of interest D (2, 2).
In order to eliminate this influence, the transmission characteristics (r-characteristics) of light intensity and image density peculiar to the electrophotographic printer G used in this relationship are set to be optimal.

In addition, here, 25 of 0 to 256 is used as digital image information.
It is quantized into 6 values (8 bits) and shows 9 cases.The differential waveform 1ζ of the transfer characteristic of light amount-image density on the printer side shown in Fig. 5 is approximated to the distribution of the dynamic characteristics shown in Fig. 2. There is.

Note that the above-mentioned spatial filter processing is performed over the entire area of the image plane read by the scanner, sequentially extracting each pixel data in a 3x3 pixel area. The spatial frequency characteristics of the Laplacian filter are shown.The horizontal axis shows the image frequency (spatial frequency) t1, the vertical axis shows the MTF value, and 17
't%. As shown in the figure, it changes depending on the value of the center value 2 of the spatial frequency characteristic line. The value of the image frequency N/2 changes depending on the sampling pitch country of the corresponding rliiB. For example, if the sampling pitch is 125 pm
Then, the frequency becomes N/2 = 1000/125×2, and in this case, the scanner ccL runs IE! The F degree is 4 pixels/mm.

Alternatively, the t-th calculation of this efficient spatial filter processing is performed by the arithmetic processing section of the system's CPU. 7 of the process that kills the arithmetic processing unit at that time
E2- is shown in FIG.

In this way, in the image control system according to the present invention, the low-frequency It changes the transmission characteristics of the components and relatively changes the transmission characteristics of the high frequency components. Then, depending on the setting of the image read by the scanners SC and C, bypass filter processing is performed dynamically in the low Ilk portion or the high 1jK portion, and 'C is a weak bypass filter for the intermediate dark portion. processing (transmission rate of DC component is high μ), reduces the loss of image information read by the scanner, and performs optimal pre-processing for the electrophotographic printer. - It becomes possible to obtain a good memorized image due to the resolution and gradation reproducibility of MVS.

Fig. 5 shows another example of the Laplacian filter. Here, the 3×3 configuration has two central values and four filter elements on the upper, lower, left and right sides for each center value, and the spatial filter processing result in this case is as follows.

-IXD (3,2 no IXD (February, March -(
2) As a result, the computation for spatial filter processing becomes easier Cζ than in the case of t using a Laplacian filter with - value in Figure 1. Note that the values of the surrounding filter elements are not limited to -1. This n is also determined by the value relative to the center value 2, and it is sufficient to fix both the center value 2 and the peripheral values. Shows frequency characteristics.

In this case as well, the value of the center value 2 of the Laplacian filter serves as a bypass filter, and the transmission of low frequency components can be changed.

More than the effects, the image processing method by Akira Honmei - ζ is read in units of #I elements by one scanner, nl and multi-value quantized, and changes the weight of the center value of the Laplacian filter according to the value of digital image information. It is designed to improve the frequency characteristics by correcting the deterioration of the frequency characteristics caused by the non-linearity of the light intensity-image density transfer characteristic of electrophotographic printers, making it possible to obtain high-quality recorded images. It has this excellent advantage.

[Brief explanation of the drawing]

FIG. 1(a) shows an example of a Laplacian filter,
The same figure shows the digital information in one pixel region extracted corresponding to the Laplacian filter. Figure 2 shows an example of the change characteristics of the weight of the center value of the Laplacian filter according to the digital image information. Characteristic diagram, Figure 3 is a diagram showing the spatial frequency characteristics of the Laplacian filter,
FIG. 4 is a flowchart of image processing according to the present invention, and FIG.
The figure shows another configuration example of the Laplacian filter, @6
Figure 87 shows the spatial frequency characteristics of the Laplacian filter, and Figure 87 shows a simplified configuration diagram of a typical laser printer.
Figure 8 shows the amount of light in an electrophotographic printer - image aIL
Figure 9 shows the transfer characteristic of the image frequency. Fig. 1 Young direction Fig. 2 Digital data acquisition □ Fig. 3 Fig. 4 Fig. 5 Fig. 6 Fig. 64

Claims (1)

[Claims] When performing spatial filter processing using a Laplacian filter on digital image information that has been read pixel by pixel by a scanner and has been quantized, the weight of the center value of the Laplacian filter is changed according to the value of the digital image information. An image processing method that takes measures.