JP3738951B2 - Image processing device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image processing apparatus that emphasizes an edge of an original image.
[0002]
[Prior art]
Conventionally, as image processing for performing filtering processing such as edge enhancement on an image, a Laplacian filter processing using a primary differential filter in the main scanning direction or a 3 × 3, 5 × 5 matrix is generally used. As processing means, 3 × 3 Laplacian filter processing is often used.
[0003]
[Problems to be solved by the invention]
However, in the Laplacian filter processing using this 3 × 3 matrix, when the edge enhancement degree is increased, steep edges that are not present in the original image are generated before and after the edges, resulting in deterioration of the processed image.
[0004]
For example, in the case of an image in which characters and pictures are mixed, it is usually necessary to emphasize the edge of the character portion, and therefore edge enhancement is performed focusing on the character portion. In this case, the image portion is processed with the same edge enhancement as the character portion, and the picture portion becomes unnatural. Therefore, there is a problem that the degree of edge enhancement cannot be increased too much.
[0005]
In order to solve such a problem, it can be considered that the coefficient of the Laplacian filter is adaptively changed according to the type of image such as a character or a picture. However, there is a problem that the processing for discriminating the image type becomes complicated, and the circuit scale becomes large because a memory or the like for holding the filter coefficient is required.
[0006]
In view of the above circumstances, an object of the present invention is to provide an image processing apparatus that can obtain a relatively good processed image without increasing the circuit scale and increasing the edge enhancement degree.
[Means for Solving the Problems]
In order to solve the above-described problem, the present invention is an image processing apparatus that edge-enhances an original image on a recording medium read by line scanning, and sets an arbitrary filter coefficient in the main scanning direction of the original image. A first differentiating means for firstly differentiating the original image in the main scanning direction; and a second differentiating means for secondarily differentiating the original image by setting arbitrary filter coefficients in the main scanning direction and the sub-scanning direction of the original image. And an inclination direction determination unit that determines whether the gradient direction of the density of the original image is positive or negative based on the output of the second differentiation unit, and a gradient of the density of the original image based on the determination result of the gradient direction determination unit. Selecting means for selectively deriving the output of the first differentiating means when the direction changes.
[0007]
The second differentiating means is preferably composed of a Laplacian filter.
[0008]
According to this configuration, the sharp edge portion of the output image unnecessarily enhanced by the second derivative is selectively replaced with the image processed by the first derivative, so that the original image before and after the edge portion of the output image is obtained. It is possible to remove extremely low density or high density areas that are not present in the image. In addition, a logic circuit for determining the gradient of the density is provided to selectively derive the image subjected to the primary differentiation process and the image subjected to the secondary differentiation process on the basis of the change in the density gradient of the original image obtained by the secondary differentiation. It can be configured simply by adding.
[0009]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
[0010]
FIG. 1 is a block diagram showing a configuration of an image processing apparatus according to an embodiment of the present invention. As shown in the figure, the image processing apparatus sets up an arbitrary filter coefficient in the main scanning direction and the sub-scanning direction of the original image and performs a 3 × 3 matrix Laplacian filter (second differentiation). Means) 100, a first-order differential filter (first differentiation means) 101 for setting an arbitrary filter coefficient in the main scanning direction of the original image and first-differentiating the original image in the main scanning direction, and an output of the Laplacian filter 100 When the inclination direction of the density of the original image changes based on the determination result of the inclination direction determination unit (inclination direction determination unit) that determines whether the inclination direction of the density of the original image is positive or negative, and the inclination direction determination unit 102 Comprises a selection circuit (selection means) for selectively deriving the output of the primary differential filter 101.
[0011]
The Laplacian filter 100 includes three image data input units, and image data scanned by one line is input to each input unit pixel by pixel. When the image data of the nth line is the target image line, the image data of the (n−1) th line is the image data of the previous line, and the image data of the (n + 1) th line is the image data of the next line.
[0012]
The Laplacian filter 100 performs edge detection and edge enhancement in the main scanning direction and the sub-scanning direction based on image data of three lines according to an arbitrarily set matrix coefficient.
[0013]
The image data of the nth line is supplied to the primary differential filter 101, and edge detection and edge enhancement in the main scanning direction are performed.
[0014]
In the image signal processed by the Laplacian filter 100, the output of the Laplacian filter changes from positive to negative or from negative to positive in a region where the gradient of the image density changes abruptly.
[0015]
The inclination direction determination unit 102 detects the image portion where the positive / negative inversion is detected , determines the positive / negative of the inclination direction of the density of the original image, and the positive / negative inversion occurs when the inclination direction of the density of the original image changes. The selection circuit 103 generates a selection signal, and the selection circuit 103 selects the output of the first-order differential filter 101. On the other hand, when the gradient direction of the density of the original image does not change and the positive / negative inversion does not occur, the output of the Laplacian filter 100 is selected.
[0016]
The Laplacian filter 100 and the first-order differential filter 101 will be described below with reference to the filter processing matrix shown in FIG. FIG. 2A shows a configuration of a Laplacian filter based on a general 3 × 3 matrix. The filter coefficient can be arbitrarily set by a controller or the like that controls the apparatus according to edge enhancement.
[0017]
Now, the density of each point in an arbitrary square area centered on a certain pixel point (i, j) in the original image is set to a new density f (i, j) using Laplacian filter processing. Thus, in the case of a 3 × 3 square area, as shown in FIG. 2B, the pixel point (i−1, j−1) to the pixel point (i + 1, j + 1) with the pixel point (i, j) as the center. In the two-dimensional image g (i, j) composed of up to nine pixels, the relationship is f (i, j) = g (i, j) * L ₄ (i, j).
[0018]
Since the Laplacian filter is a filter used for extracting an edge portion, it is necessary to apply a threshold value operation in order to actually use the edge enhancement process. Here, the filter coefficient shown in FIG.
[0019]
FIG. 2D shows the configuration of the first-order differential filter 101. The filter coefficient can be arbitrarily set by a controller or the like that controls the apparatus according to edge enhancement.
[0020]
Further, the filtering process will be described with reference to FIG. FIG. 3A shows an example of the input original image. The vertical axis represents the image density, and the horizontal axis represents the main scanning direction of the target image line. FIG. 3B shows the result of filtering the input original image by the Laplacian filter 100 shown in FIG. 1, and FIG. 3C shows the result of FIG. The result of having performed the filter process by the 1 differential filter 101 shown in FIG.
[0021]
FIG. 3D shows the positive / negative state of the inclination direction determined by the inclination direction determination unit 102 shown in FIG. 1 based on the filter processing by the Laplacian filter 100. In the figure, the vertical axis indicates the slope level, and the higher the level, the steeper the slope.
[0022]
The tilt direction determination unit 102 sends a selection signal to the selection circuit 103 shown in FIG. 1 so as to select the output of the first-order differential filter 101 at a change point where the tilt direction changes from positive to negative or from negative to positive. appear.
[0023]
FIG. 3E shows an edge enhanced image output from the selection circuit 103. As shown in the figure, it is confirmed that the edge-enhanced image does not include an image portion having a steep density gradient before and after the edge that does not exist in the original image (FIG. 3A).
[0024]
As described above, a steep gradient portion before and after the edge that does not exist in the original image generated by the secondary differential filter processing having a high edge enhancement degree is based on the sign of the density gradient of the image obtained by the secondary differential filter processing. By identifying and replacing this portion with an image processed with a first-order differential filter having a low edge enhancement degree, it is possible to perform edge enhancement faithfully to the original image. Further, it is not necessary to newly provide a circuit for identifying a steep gradient portion before and after the edge, and since the output generated by the secondary differential filter processing is used, the circuit configuration does not increase. Furthermore, in the inclination direction determination based on the secondary differential filter process, since it is a simple determination that only detects a change point where the sign changes, high-speed determination is possible. Therefore, high-speed edge enhancement is possible. In particular, an image processing apparatus that reads an image in units of lines can sequentially process the read line images, so there is no need to add a memory or the like for holding the images.
[0025]
【The invention's effect】
As described above, the image processing apparatus of the present invention is an image processing apparatus that edge-enhances an original image on a recording medium read by line scanning, and sets an arbitrary filter coefficient in the main scanning direction of the original image. A first differentiating means for firstly differentiating the original image in the main scanning direction; and a second differentiating means for secondarily differentiating the original image by setting arbitrary filter coefficients in the main scanning direction and the sub-scanning direction of the original image. Differentiating means, inclination direction determining means for determining the sign of the inclination direction of the density of the original image based on the output of the second differentiating means, and density of the original image based on the determination result of the inclination direction determining means When the inclination direction of the first change means changes, the selection means for selecting and deriving the output of the first differentiation means is provided, so that the circuit scale is not increased and the edge enhancement degree is relatively good. Get processed image Door can be.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of an image processing apparatus according to an embodiment of the present invention.
FIG. 2 is a diagram showing a filter processing matrix according to the present invention.
FIG. 3 is a view showing a processing result by a filter processing according to the present invention.
[Explanation of symbols]
100 Laplacian Filter 101 Primary Differential Filter 102 Inclination Direction Determination Unit 103 Selection Circuit

Claims (2)

An image processing apparatus for edge enhancement of an original image on a recording medium read by line scanning,
First differentiation means for setting an arbitrary filter coefficient in the main scanning direction of the original image and firstly differentiating the original image in the main scanning direction;
Second differentiation means for secondarily differentiating the original image by setting arbitrary filter coefficients in the main scanning direction and the sub-scanning direction of the original image;
A tilt direction determining unit that determines whether the density of the original image has a gradient direction based on the output of the second differentiating unit;
Selection means for selectively deriving the output of the first differentiating means when the inclination direction of the density of the original image changes based on the determination result of the inclination direction determining means;
An image processing apparatus comprising: The image processing apparatus according to claim 1, wherein the second differentiating unit is a Laplacian filter.

Images