JP2000253235A - Picture processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a picture processing method, by which the picture of low resolution is converted into that of high resolution, capable of suppressing the increase of circuit scale and the increase of cost, easily securing the similarity of a processing and improving the quality of the picture subjected to resolution conversion. SOLUTION: When a regard picture element PT is positioned close to one of peripheral picture elements PA, PB, PC and PD, the picture element value of the peripheral picture element which the regard picture element PT approaches is set to be the picture element of the regard picture element PT. When the regard picture element PT is not positioned close to any peripheral picture elements PA, PB, PC and PD, the result of a product-sum operation using the picture element values of the peripheral picture elements PA, PB, PC and PD is set to be the picture element value of the regard picture element PT.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an image processing method for converting a low-resolution image into a high-resolution image.

[0002]

2. Description of the Related Art Various image processing methods for converting a low-resolution image into a high-resolution image have been proposed.
For example, as shown in FIG. 4A, original pixels (hereinafter, referred to as “surrounding pixels”) P _A , P _B , surrounding the pixel of interest _PT .
P _C, nearest the interpolation method and, FIG to the pixel of interest P _T pixel value of the nearest pixel to the pixel of interest P _T of P _D (b)
, The interval between the original pixels is d, and the pixel values of the surrounding pixels P _A , P _B , P _C , and P _D are A, B, C, and D, respectively.
And then, also, the horizontal direction the pixel of interest P _T from surrounding pixels P _A, in the vertical direction respectively distances i, assuming that separated by j, the value T of the interpolated prime _{P T, T = {(d} -i) (d−j) A + i (d
A bilinear interpolation method is known in which the result of the product-sum operation using the pixel values of the surrounding pixels is used as the pixel value of the pixel of interest as −j) B + (di) jC + ijD｝ / d ² .

[0003]

The nearest neighbor interpolation method and the bilinear interpolation method have the advantage that they can be realized with a simple configuration, but have the following problems. First, in the closest interpolation method, blocks are visually conspicuous, and the image quality is not good. Further, in the bilinear interpolation method, a screen with smoothing is obtained, but sharp portions and edge portions of characters and figures are blurred.

For this reason, in practice, resolution conversion is often performed by a more complicated method. For example, more original pixel values are used to determine the pixel value of the pixel of interest, or contour correction component data is generated and added to the pixel value obtained by performing resolution conversion using a relatively simple method. Or there is a way.

However, in order to realize these methods, the configuration becomes complicated, so that there is a problem that the circuit scale of the device for performing the resolution conversion increases and the cost increases. Further, since the processing is complicated, there is a problem that it is difficult to secure the synchronization of the processing.

Accordingly, the present invention is an image processing method for converting a low-resolution image into a high-resolution image, which suppresses an increase in hardware circuit scale and cost, and facilitates securing of processing simultaneity. It is another object of the present invention to provide an image processing method in which the image quality of an image after resolution conversion is improved.

[0007]

In order to achieve the above object, according to the first aspect of the present invention, in an image processing method for converting a low-resolution image to a high-resolution image, an attempt is made to obtain the pixel value. Depending on the positional relationship between the pixel of interest that is a pixel and a peripheral pixel that is a pixel located around the pixel of interest among the original pixels that are the pixels of the original image that is the resolution conversion target, One of the pixel value of the pixel closest to the target pixel and the result of the product-sum operation using the pixel values of the surrounding pixels is set as the pixel value of the target pixel.

According to the above method, the pixel values of the original pixels which are close to each other in a range where the block-shaped pattern is not conspicuous are substituted. Then, linear interpolation can be performed.

Also, based on the result of a predetermined operation using the pixel values of the surrounding pixels, it is determined whether the pixel value of the target pixel is obtained by the above method or the pixel value of the target pixel is obtained by another method. Switching may be performed.

For example, it is possible to determine whether an image is an edge component, a character or a figure, based on the magnitude of the difference between the pixel values of adjacent peripheral pixels.
Whether the interpolation is performed by the image processing method according to the first aspect or the interpolation is performed by another method is switched according to the attribute of the image.

[0011]

Embodiments of the present invention will be described below with reference to the drawings. In the image processing method according to the first embodiment of the present invention, the pixel value of the target pixel is obtained by a method corresponding to the positional relationship between the target pixel and surrounding pixels. Specifically, as shown in FIG. 1, with respect to the target pixel _PT and the four surrounding pixels P _A , P _B , P _C , and P _D ,
Each interval between the two peripheral pixels adjacent in the vertical direction d _x, d _y, i laterally target pixel P _T from surrounding pixels P _A,
When the vertical direction is referred to as being j by each remote location to determine the value T of the target pixel P _T as follows.

[0012] (1) i <time t _x and j <t _y (i.e., when the pixel of interest P _T is located in the region R1) is, T
= A. (2) when _{t x ≦ i ≦ d x -t} x and j <t _y (i.e., when the pixel of interest P _T is positioned in the region R2) on _{the, T = {(d x -t} x -i) A + (i−t _x ) B｝ / (d _x −2t
_x ). ₍₃₎ when the d _x -t x <i and j <t _y (i.e., when the pixel of interest P _T is positioned in the region R3), the the T = B. (4) i <time t _x and _{t y ≦ j ≦ d y -t} y ( i.e., when the pixel of interest P _T is positioned in the region R4) _{The, T = {(d y -t} y -j) A + (j− _ty ) C｝ / ( _dy− 2)
t _y ). (5) when _{t x ≦ i ≦ d x -t} x and _{t y ≦ j ≦ d y -t} y ( i.e., when the pixel of interest P _T is located in the region R5)
_{The, T = (K 1 K 2} A + K 2 K 3 B + K 1 K 4 C + K 3 K 4 D) / (d x
−2t _x ) (d _y −2t _y ). Where K ₁ = d _x −t _x −i, K ₂ = d _y −t _y −
j, is _{K 3 = i-t x,} K 4 = j-t y. _{(6) d x -t x <} i and when _{t y ≦ j ≦ d y -t} y ( i.e., when the pixel of interest P _T is located in a region R6) _{To, T = {(d y -t} y −j) B + (j− _ty ) D｝ / ( _dy− 2)
t _y ). (7) When i <t _x and d−t _y <j (that is, when the target pixel _PT is located in the region R7), T = C. (8) t when _x ≦ i ≦ d _x -t _x and d _y -t _y <j (i.e., when the pixel of interest P _T is positioned in the region R8) _{To, T = {(d x -t} x −i) C + (i−t _x ) D｝ / (d _x −2)
t _x ). (9) When d _x -t _x <i and d _y -t _y <j (i.e., when the pixel of interest P _T is positioned in the region R9) The, T
= D.

[0013] Here, the horizontal direction, the vertical direction of the considered separately, for example, two surrounding pixels P _A in FIG. 1, when viewed by focusing on interpolation of the only transverse with P _B, the pixel of interest P _T the pixel value T, i <t _x sometimes, when _{T = a t x ≦ i ≦} d x -t x _{is, T = {(d x -t} x -i) a + (i-t x) B} / (d _x -2
t _x ) d _x −t _x <i, T = B, and the target pixel P _T
Is represented by T = αA + βB, i and α,
i and β have the relationships indicated by solid lines in FIGS. 2A and 2B, respectively. Note that i and α, i and β by the bilinear interpolation method
Is shown by a broken line in FIG.

That is, in the image processing method of the first embodiment, the interpolation method is switched according to the positional relationship between the target pixel and the surrounding pixels. Specifically, if the target pixel is close to the surrounding pixel, the pixel value of the surrounding pixel is set to the pixel value of the target pixel. On the other hand, if the target pixel is not close to the surrounding pixel, the pixel value of the peripheral pixel is set. Is used as the pixel value of the target pixel. Note that the distance t _x in the horizontal direction
As each threshold distance t _y in the vertical direction, the pixel of interest is determined whether or not close to the surrounding pixels.

Here, for example, small characters and figures used in computer screen output often have a small number of pixels and thin lines. If this is enlarged at a low magnification (about 1 to 3 times) by a conventional method, the original form of the original image may be lost, which may make it difficult to read. In particular, when the bilinear interpolation method is used, an oblique straight line having a width of one pixel is easily affected.

On the other hand, in the image processing method of the first embodiment, the distances t _x and t _y are set to, for example, 0.25 d _x , 0.2
Setting each to a value such as 5d _y simplifies the circuit configuration. Then, as the pixel value of the pixel of interest, the pixel value of the original pixel that is close in a range where the block-shaped pattern is not conspicuous is substituted. Since linear interpolation is performed after emphasizing, when sharp image output of characters and figures is required, the image after resolution conversion is less blurred compared to the bilinear interpolation method. Become. Also, the blocks are less noticeable compared to the nearest interpolation method. Therefore, it can be said that the image processing method of the first embodiment is a method particularly suitable for performing resolution conversion by enlarging an image composed of fine characters and graphics at a low magnification.

In addition, there has been a conventional method of performing contour correction using a filter such as Laplacian. However, in this method, an overflow or an underflow may occur in a calculation, and it is necessary to provide a circuit limiter. On the other hand, in the image processing method of the first embodiment, such a problem does not occur, and the image processing method can be realized with a circuit scale substantially equal to that of the bilinear interpolation method. In addition, it is easy to ensure the simultaneity of the processing.

Next, an image processing method according to a second embodiment of the present invention will be described. FIG. 3 is a block diagram of an image processing apparatus that realizes the image processing method according to the second embodiment. Each block shown in FIG.

Reference numeral 1 denotes an input terminal to which a pixel value of an original pixel constituting an image to be converted is input. Reference numeral 2 denotes a linear interpolation circuit that performs interpolation by a bilinear interpolation method based on the pixel values input from the input terminal 1. Reference numeral 3 denotes a non-linear interpolation circuit for performing interpolation by the method of the first embodiment based on the pixel value input from the input terminal 1. Reference numeral 4 denotes the input of the output of the linear interpolation circuit 2 and the output of the non-linear interpolation circuit 3,
It is a selector for selecting either one.

Reference numeral 5 denotes an adjacent pixel difference circuit for obtaining an adjacent pixel difference which is a difference between pixel values of original pixels surrounding the target pixel. 6
Compares the adjacent pixel difference obtained by the adjacent pixel difference circuit 5 with the threshold value input from the threshold value input terminal 7, and as a result, if the adjacent pixel difference is smaller than the threshold value, the linear interpolation circuit 2
Is a comparator that controls the selector 4 so as to select the output of the nonlinear interpolation circuit 3 when the adjacent pixel difference is larger than the threshold value. 8 is a selector 4
Is an output terminal for outputting the pixel value selected by.

With the above configuration, interpolation is performed by the image processing method of the first embodiment on a portion where the difference between adjacent pixels is larger than the threshold value, and on a portion where the difference between adjacent pixels is smaller than the threshold value. Interpolation is performed by a linear interpolation method.

Accordingly, the difference in pixel value between adjacent pixels in an edge component or a character / figure becomes large. Interpolation is performed by the method of the first embodiment, which functions to preserve the components of the above, while in other cases interpolation is performed by a bilinear interpolation method with smoothing.

As described above, in the method of the second embodiment, the interpolation is performed in a method suitable for the attribute of the image (whether or not it is an edge component or a character or a figure). Can be improved. The image quality can be adjusted by controlling the threshold value input to the threshold value input terminal 7.

Since the method of the first embodiment is highly compatible with the bilinear interpolation method in terms of circuitry, the linear interpolation circuit 2 for performing interpolation by the bilinear interpolation method and the linear interpolation circuit 2 of the first embodiment are described. Many parts of the circuit configuration can be shared with the non-linear interpolation circuit 3 that performs interpolation by the method, so that an increase in the circuit scale can be suppressed as much as possible.

In the image processing method according to the second embodiment,
Depending on the magnitude of the difference between adjacent pixels, interpolation is performed by a bilinear interpolation method instead of the image processing method of the first embodiment, but other interpolation is performed according to the attribute of the image to be resolution-converted. An interpolation method may be used.

In the image processing method according to the second embodiment, the magnitude of the difference between adjacent pixels is compared with one threshold value, so that the image processing method according to the first embodiment can be compared with another interpolation method. The method is switched to one of the two interpolation methods. However, by comparing the magnitude of the difference between adjacent pixels with a plurality of threshold values, the image processing method of the first embodiment and the other plurality of interpolation methods are compared. May be switched to any one of the three or more interpolation methods.

Further, in the image processing method of the second embodiment, the switching of the interpolation method is performed based on the magnitude of the difference between the adjacent pixels. The calculation may be performed based on another calculation result using the value. What kind of calculation should be performed may be determined according to the attribute of the image whose resolution is to be converted.

[0028]

As described above, according to the image processing method of the present invention, when converting a low-resolution image into a high-resolution image, it is possible to suppress an increase in circuit scale and cost and to simultaneously perform processing. After ensuring the
The image quality of the image after the resolution conversion can be improved.
The image processing method according to the first aspect is particularly suitable for the case where resolution conversion is performed by enlarging an image composed of fine characters and graphics at a low magnification. In particular, the image processing method according to the second embodiment can improve an image after resolution conversion for an image with more attributes.

[Brief description of the drawings]

FIG. 1 is a diagram showing a positional relationship between a target pixel and surrounding pixels.

FIG. 2 is a diagram illustrating a relationship between a coefficient and a position of a pixel of interest when the image processing method according to the first embodiment is represented by a product-sum operation when interpolation is considered in a one-dimensional direction.

FIG. 3 is a block diagram of an image processing apparatus that realizes an image processing method according to a second embodiment of the present invention.

FIG. 4 is a diagram illustrating a nearest interpolation method and a bilinear interpolation method.

[Explanation of symbols]

1 input terminal 2 the linear interpolation circuit 3 nonlinear interpolation circuit 4 selector 5 adjacent pixel difference circuit 6 comparator 7 threshold input terminal 8 an output terminal P _T target pixel _{P A, P B, P C} , P D surrounding pixels

Claims (1)

[Claims] In an image processing method for converting a low-resolution image into a high-resolution image, a pixel of interest whose pixel value is to be obtained and an original pixel which is a pixel of an original image whose resolution is to be converted are included. The pixel value of the pixel closest to the target pixel among the peripheral pixels and the pixel value of the peripheral pixel were used in accordance with the positional relationship between the target pixel and the peripheral pixel that is a pixel located around the target pixel. An image processing method, wherein one of the result of the product-sum operation is set as the pixel value of the target pixel.