JPH0877350A - Image processor - Google Patents

Abstract

PURPOSE: To provide the image processor for performing smoothing and edge emphasis corresponding to the image kind such as a character or a photograph in the process for processing the image data read by an image reader such as an image scanner. CONSTITUTION: The output of a low-pass filter 2 for performing smoothing and the output provided through a high-pass filter 3 and a noise removing/edge emphasizing part 4 are synthesized by a synthesizing circuit 5. The noise removing/edge emphasizing part 4 is composed of plural look up tables(LUT 1-n) and the most suitable look up table is selected out of them and used for each noticed picture element to be processed. The selection is performed corresponding to the density difference between the density (such as an average value from an averaging circuit 11 in the case of a photograph or picture image, a maximum value or minimum value from a maximum/minimum detecting circuit 12 in the case of a character image) found with respect to a filter area and the density of the noticed picture element.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for performing filter processing according to image types such as character images and photographic images in the process of processing image data read by an image reading apparatus such as an image scanner. It is about.

[0002]

2. Description of the Related Art An image to be read includes an image containing only characters, an image containing only photographs and pictures, or an image in which these are mixed in one page. If these images are printed or displayed on the display with the image data just read by the image scanner, the reproducibility of the images is poor.
Therefore, smoothing processing (smoothing) is performed on a photograph or picture image, and edge enhancement processing is performed on a character image. These processes are performed by the image processing device having a two-dimensional filter as a main component.

The position occupied by such an image processing apparatus in the overall configuration will be described by taking a color copying machine as an example. FIG. 4 is a diagram showing an outline of the overall configuration of the color copying machine. The image processing device of the present invention corresponds to 27 parts. The reflected light from the document is read by the image sensor 20, converted into a digital value by the AD conversion unit 21, and separated by the line memory 22 into RGB (red, green, black) components. From the image sensor 20 to the line memory 22, the RGB components of the image data are sent dot-sequentially.

The RGB components sent from the line memory 22 to the color conversion unit 23 are converted into YMC (yellow, magenta, cyan) components. Under color removal / black generation unit 24 (Under color removal ... UCR: Under Color Removal)
Then, the under color is removed and the black color (BK) is generated. Then, the BK, C, M, and Y data are transferred to the TRC correction unit 25 (Tonor Reproduction Curv.
e) sent to.

The enlargement / reduction processing unit 26 enlarges or reduces the image as necessary. Next, the image processing device 27 performs smoothing or edge enhancement, and the halftone generator 28 performs dither method and error diffusion processing. The image data processed as described above is IOT I / F 29 (Image Ou
tput terminal interface) and sent to the IOT.

By the way, even if a two-dimensional filter is used for smoothing and edge enhancement, filters having different characteristics depending on the type of original image (in other words, filters having different filter coefficient values).
Need to use. In the case of an image in which text and pictures (pictures) are mixed, it may be processed with a single filter.
Good results are not obtained. If the characteristics of the filter are suitable for the processing of character images, the processing of the photo part will not be performed properly, and if the characteristics of the filter are suitable for the processing of the photographic image, the processing of the character part will be performed. Is not done properly.

Therefore, the frequency characteristic of the document is analyzed by TI separation (text / image separation), and the characteristic of the filter to be used is changed based on the analysis result. However, in order to analyze the frequency characteristics of the manuscript, a filter having various frequency characteristics is separately required, and a circuit such as a DSP (digital signal processor) is required, so that the circuit configuration is complicated. Becomes A filter having various frequency characteristics is expensive.

An image processing apparatus has been considered in which the circuit configuration does not become so complicated and the cost is not increased because a filter having various frequency characteristics is not used. It has a low-pass filter suitable for smoothing and a high-pass filter suitable for edge enhancement, passes the read image data through these filters separately, and removes noise and edge for the image data coming through the high-pass filter. The image processing apparatus is configured to perform emphasis processing and combine the results later.

FIG. 5 is a block diagram of such a conventional image processing apparatus. In FIG. 5, 1 is a line memory unit, 2 is a low pass filter, 3 is a high pass filter, 4 is a noise removing / edge enhancing unit, and 5 is a synthesizing circuit. In this example, the line memory unit 1 is configured by a line memory for 5 lines. Therefore, the low-pass filter 2 and the high-pass filter 3
Is a filter that processes image data in a predetermined filter area (for example, a 5 × 5 matrix area centered on a pixel of interest), and these have a predetermined number of adjacent lines (eg, 5 lines). Image data of is input all at once.

FIG. 3A shows the characteristic of the low-pass filter 2, and FIG. 3B shows the characteristic of the high-pass filter 3. The low pass filter 2 outputs a low frequency component as a large value. The high-pass filter 3 outputs a frequency component belonging to a constant high frequency bandwidth as a large value. Therefore, from the low-pass filter 2, an output is obtained where the portion where the density changes gently (the portion such as a photograph (picture)) is smoothed. From the high-pass filter 3, an output of a portion where the density changes abruptly, such as an edge portion, is obtained.

The output that has passed through the high-pass filter 3 contains some noise. Therefore, the noise removal / edge enhancement unit 4 is provided in the next stage of the high pass filter 3 to remove noise and edge enhance the output of the high pass filter 3. Noise removal / edge enhancement unit 4
Is specifically composed of a lookup table.
In this lookup table, the high pass filter 3
The table is accessed using the input from the address as an address, and the value stored in the accessed address is read and output.

When the output of the low-pass filter 2 and the output of the noise removal / edge enhancement unit 4 are combined by the combining circuit 5, even if the image includes characters and a picture (picture), the picture (picture) part is smoothed. As a result, an image in which the character portion is edge-enhanced is obtained.

Incidentally, as a conventional document relating to such an image processing apparatus, for example, JP-A-3-32160,
There are JP-A-3-88478 and JP-A-3-185574.

[0014]

In the conventional image processing apparatus as shown in FIG. 5, the contents of the look-up table constituting the noise removing / edge enhancing section 4 are fixed, and different types of originals are recorded. The lookup table having the same content is also applied to the areas that are different from each other or the areas of the same original document having different original states. Therefore, there is a problem in that some areas may not be properly subjected to noise removal and edge enhancement depending on the document conditions. An object of the present invention is to solve such a problem.

[0015]

In order to solve the above problems, in the present invention, a low-pass filter and a high-pass filter to which image data of a predetermined filter area is separately input, and noise of the output of the high-pass filter are removed. In the image processing apparatus, which comprises a noise removal / edge enhancement means for performing edge enhancement, and a synthesis circuit that synthesizes the output of the low-pass filter and the output of the noise removal / edge enhancement means, A density difference calculating unit configured by a plurality of lookup tables that are selectively used and calculating a difference between the density of the pixel of interest and the density obtained for the filter region of the pixel of interest,
Threshold storage means for storing a plurality of thresholds set corresponding to the plurality of look-up tables, and a density difference level for detecting a density difference level by comparing the calculated density difference and the threshold. With a detection circuit, the detection signal from the density difference level detection circuit
It was decided to decide the lookup table to be selectively used by the noise removal / edge enhancement means.

The density difference calculating means includes an averaging circuit for obtaining an average value of the densities of the filter area, a maximum / minimum detecting circuit for obtaining a maximum value and a minimum value of the density of the filter area, and the averaging circuit. A selection circuit for selecting whether it is a maximum / minimum detection circuit and a subtractor for calculating a density difference between the density of the pixel of interest and the average value or a density difference between the density of the pixel of interest and the maximum value or the minimum value Can be

The threshold value of the threshold storage means can be changed according to the type of the original image.

[0018]

[Operation] The noise removal / edge enhancement unit connected to the subsequent stage of the high-pass filter is configured with multiple lookup tables instead of a single lookup table, and the most appropriate look is obtained for each target pixel to be processed. Select and use the up table. As a result, it is possible to finely adjust the noise situation and edge enhancement according to the type of image without using a high-cost filter having various frequency characteristics or using a DSP or the like to complicate the circuit configuration. Become.

[0019]

Embodiments of the present invention will now be described in detail with reference to the drawings. FIG. 1 is a block diagram of the image processing apparatus of the present invention. Reference numerals correspond to those of FIG. 5, 6 is an operation panel, 7 is a CPU (central processing unit), 8 is a density difference calculation unit, 9 is a delay circuit, 10 is a selection circuit, 11 is an averaging circuit, and 12 is A maximum / minimum detection circuit, 13 is a subtractor, 14 is a density difference level detection circuit, and 15 is a threshold storage unit. The selection circuit 10 is a circuit that selects either the averaging circuit 11 or the maximum / minimum detection circuit 12.

Since the parts having the same reference numerals as those in FIG. 5 operate in the same manner, the description thereof will be omitted. In terms of configuration, the first difference from the conventional example of FIG. 5 is that the noise removal / edge enhancement unit 4 is composed of a plurality of lookup tables LUT1 to LUTn, and one of them is selected and used. That is the point. The second difference is that the density difference calculation unit 8 and the density difference level detection circuit 14 are configured to generate the selection signal.
Also, the threshold storage unit 15 is provided.

FIG. 2 is a diagram showing the input / output characteristics of the noise removal / edge enhancement unit. One characteristic curve corresponds to one look-up table. Here, in order to avoid complication of the drawing, only three characteristic curves a, b, and c are drawn, but in reality, the number of characteristic curves is the same as the number of lookup tables installed.

In the range W from -A ₁ to A ₁ of the input,
Although the output is set to 0, noise removal is realized by this. That is, an input having a size belonging to the range W is regarded as a noise component, and when such an input comes in, the output becomes 0. Further, the characteristic curve located in the + arrow direction has a higher degree of edge enhancement, and the characteristic curve located in a − arrow direction has a lower degree of edge enhancement.
For example, the characteristic curve a has a higher degree of edge enhancement than the characteristic curve c. It can be understood from the fact that even with the same input, the characteristic curve a produces a larger output.

The image reproducibility can be made good by selecting the characteristic curve of the edge emphasis degree which seems to be appropriate for the image data to be processed. A signal for selecting an appropriate degree is generated by the density difference calculation unit 8, the density difference level detection circuit 14 and the threshold storage unit 15, which will be described below.

The density difference calculating section 8 is a section for calculating the density difference between the density of the pixel of interest and the "density calculated for the filter area". The "density obtained for the filter area" is specifically either the "average value" or the "maximum value" and "minimum value" of the density of the filter area of the pixel of interest, and whichever is selected. Whether or not the operator considers the type of original image (whether to improve the reproducibility of the photograph (picture) portion or to make the character portion particularly clear) is determined.

That is, when it is desired to improve the reproducibility of the portion of the photograph (picture) even if the original is a photograph (picture), or if the original is mixed with text, the area within the filter area is set. The “average value” of the image data is used as the partner for calculating the density difference. Therefore, the averaging circuit 11 is selected by the selection circuit 10. The selection is made by an operator operating the operation panel 6 to instruct the CPU 7 and the CPU 7 sends a selection signal to the selection circuit 10.

The averaging circuit 11 takes the average value of the image data values in the filter area and outputs it. Taking the average value means that even if there is a prominently large value or a prominently small value in the filter area, we do not pay attention to those values, but rather place importance on the density of the entire area. That is. In a photograph (picture), the density usually does not change abruptly and usually changes gradually. Therefore, it is preferable to process the density in comparison with the average density value of the entire region. Therefore, evaluate the area around the pixel of interest with an average value,
A density difference between the pixel and the target pixel that has passed through the delay circuit 9 is calculated, and edge enhancement is performed according to the density difference.

Even if the manuscript is text or mixed with a photograph (picture), if it is desired to reproduce the text part clearly and clearly, then it is within the filter area. The largest density difference from the image data of (1) is calculated and used as a clue to determine the degree of edge enhancement. The maximum density difference is either the difference between the maximum density of the filter area and the minimum density. Therefore, the "maximum value" and the "minimum value" are the opponents for the density difference calculation. In this case, the operator operates the operation panel 6 to select to use the maximum / minimum detection circuit 12.

The maximum / minimum detection circuit 12 detects the maximum value and the minimum value of the image data in the filter area and sends it to the subtractor 13. The subtractor 13 calculates the difference between the density and the maximum value of the target pixel that has passed through the delay circuit 9 and the difference between the minimum value and the larger value, and uses the larger one. In the case of a character image, the density changes abruptly at the edge portion of the character, so edge enhancement without overlooking it will make the character clear and sharp. Therefore, the maximum density difference in the filter area is obtained, and this is used as a clue to determine the degree of edge enhancement.

In the delay circuit 9, the value of the pixel of interest is subtracted by the subtractor 13.
It is for adjusting the timing to be input to.
The other party who obtains the density difference from the target pixel is the output of the averaging circuit 11 or the maximum / minimum detection circuit 12, but the target pixel is also input in accordance with the timing when those outputs are input to the subtractor 13. Adjust the timing.

The threshold storage unit 15 has a number of thresholds (T1 to T) equal to the number of lookup tables mounted in the noise removal / edge enhancement unit 4.
n) is set. To set the threshold, the operator operates the operation panel 6 to send a command to the CPU 7, and C
This is performed by the PU 7 sending a setting signal to the threshold storage unit 15. The number of lookup tables is n
If it is, the number of thresholds is also n. The threshold value can be variously changed depending on the type of document.

The density difference level detection circuit 14 includes a subtracter 13
The density difference calculated in step 3 is compared with each threshold in the threshold storage unit 15 to detect which threshold the density difference exceeds, and the look-up corresponding to the exceeded threshold is detected. A selection signal is sent to the noise removal / edge enhancement unit 4 so as to select the up table.
For example, when the density difference exceeds the threshold T2 but does not exceed T3 above it, a signal for selecting the lookup table LUT2 corresponding to the threshold T2 is sent to the noise removal / edge emphasis unit 4. send.

The subtractor 13 calculates the density difference for each pixel of interest, and the selection signal from the density difference level detection circuit 14 is also issued for each pixel of interest. An appropriate table is selected for each pixel of interest. In the synthesis circuit 5,
Since the output from the low-pass filter 2 and the output from the high-pass filter 3 are noise-removed and edge-emphasized by an appropriate lookup table, the combined image has good reproducibility.

[0033]

As described above, according to the image processing apparatus of the present invention, the noise removing / edge enhancing unit connected to the subsequent stage of the high pass filter is not a single lookup table but a plurality of lookup tables. For each pixel of interest to be constructed and processed, the most appropriate one is selected and used. Therefore, the noise situation and the edge enhancement are finely adjusted according to the type of the image and the situation of the original document without using a high-cost filter having various frequency characteristics or using a circuit having a complicated configuration including a DSP. I was able to do it.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of an image processing apparatus of the present invention.

FIG. 2 is a diagram showing input / output characteristics of a noise removal / edge enhancement unit.

FIG. 3 is a diagram showing characteristics of a low-pass filter and a high-pass filter.

FIG. 4 is a diagram showing an outline of an overall configuration of a color copying machine.

FIG. 5 is a block configuration diagram of a conventional image processing apparatus.

[Explanation of symbols]

1 ... Line memory unit, 2 ... Low-pass filter, 3 ... High-pass filter, 4 ... Noise removal / edge enhancement unit, 5
... Synthesis circuit, 6 ... Operation panel, 7 ... CPU, 8 ... Density difference calculation unit, 9 ... Delay circuit, 10 ... Selection circuit, 11 ... Averaging circuit, 12 ... Maximum / minimum detection circuit, 13 ... Subtractor, 14 ...
Density difference level detection circuit, 15 ... Threshold storage unit

Claims (3)

[Claims] 1. A low-pass filter and a high-pass filter to which image data of a predetermined filter area is separately input, noise removal / edge enhancement means for removing noise from the output of the high-pass filter and edge enhancement, and the low-pass filter. In an image processing device comprising a synthesizing circuit for synthesizing an output of a filter and an output of the noise removing / edge enhancing means, the noise removing / edge enhancing means is constituted by a plurality of lookup tables which are selectively used. Density difference calculating means for calculating the difference between the density of a pixel and the density obtained for the filter area of the pixel of interest, and a threshold storing a plurality of thresholds set corresponding to the plurality of look-up tables. The density difference level is compared by comparing the calculated density difference and the threshold with the hold storage means. And detect the density difference level detecting circuit provided by the detection signal from the density difference level detecting circuit, an image processing apparatus characterized by determining a lookup table for selecting used in the noise removal edge enhancement means. 2. The density difference calculating means comprises an averaging circuit for obtaining an average value of densities in a filter area, a maximum / minimum detecting circuit for obtaining maximum and minimum values of densities in the filter area, and the averaging circuit or the maximum. And a subtracter for selecting a density difference between the density of the pixel of interest and the average value or a density difference between the density of the pixel of interest and the maximum value or the minimum value. The image processing apparatus according to claim 1, wherein: 3. The image processing apparatus according to claim 1, wherein the threshold value of the threshold storage means can be changed according to the type of original image.