KR100949137B1 - Image interpolation device and method and computer readable recording medium recording the method - Google Patents

Abstract

The present invention relates to an image interpolation apparatus, a method thereof, and a computer-readable recording medium on which the method is recorded. An image interpolation apparatus according to the present invention, comprising: an edge direction detector for detecting an edge direction of a current pixel; And a pixel interpolation unit configured to calculate a value of the current pixel by assigning a distance weight to the distance between the current pixel and the pixel in the edge direction and the distance between the current pixel and the pixel in the vertical direction.

Accordingly, the performance can be improved by introducing a new distance weighting concept based on an interpolation method considering a regional pattern.

Image interpolation, distance weights, interlacing

Description

A video interpolation device and a method thereof and a computer-readable recording medium recording the method thereon {APPARATUS FOR VIDEO INTERPOLATION, METHOD THEREOF AND COMPUTER RECORDABLE MEDIUM STORING THE METHOD}

The present invention relates to an image interpolation apparatus, a method thereof, and a computer-readable recording medium on which the method is recorded.

Interlacing schemes are widely used in many TV systems, including NTSC, PAL, and SECAM, because they efficiently use limited bandwidth. This interlaced scanning method has the advantage of doubling the frame rate while using the same bandwidth, but due to the nature of the scanning method, there is a flickering between the screens, It has the disadvantage that visual deterioration occurs like twitter phenomenon.

To overcome this problem, various interlacing injection methods have been proposed. These methods can be broadly divided into inter-field interlaced scanning using inter-field prediction and intra-field scanning using only information in one field.

The inter-field interlaced scanning method is classified into a method of adaptively filtering and applying motion and a method using motion compensation. The inter-field interlaced scanning method generally yields acceptable results if the motion information is reliable. However, the inter-field interlaced scanning method has a disadvantage of not only good performance if the motion information is reliable but also very complicated in hardware implementation.

Intra-field interlaced scanning methods are classified into approaches using various spatial filters and approaches considering edge orientation. Since the edge of the image and its orientation are very sensitive to human vision, blurring is likely to occur when interlaced scanning is performed without considering such information. Edge-based line averaging (ELA) algorithm is the most widely used method of intra-field interlaced scanning in consideration of the directionality of such an image. The ELA algorithm is widely used due to its simple calculation and ease of implementation, but has a disadvantage in that it is sensitive to small changes in pixel values. Many techniques have been proposed to compensate for this drawback.

On the other hand, various methods have also been tried in the method of performing interpolation after edge direction determination. The simplest method is a linear interpolation method that linearly interpolates two pixels in the edge direction. However, since this method does not consider the regional pattern, the image blurring phenomenon is likely to occur. In order to solve this problem, a method of considering a local pattern of a plurality of pixels in an edge direction has been proposed.

Accordingly, it is an object of the present invention to provide an image interpolation apparatus adopting a new distance weight concept based on an interpolation method considering a regional pattern, a method thereof, and a computer-readable recording medium on which the method is recorded.

According to an aspect of the present invention, there is provided an image interpolation apparatus, comprising: an edge direction detection unit for detecting an edge direction of a current pixel; And a pixel interpolation unit configured to calculate a value of the current pixel by assigning a distance weight to the distance between the current pixel and the pixel in the edge direction and the distance between the current pixel and the pixel in the vertical direction. Can be achieved by an image interpolation device.

The pixel interpolation unit may include an edge interpolation calculation unit configured to calculate an edge interpolation value based on pixels in the edge direction; A vertical interpolation calculator configured to calculate a vertical interpolation value based on the pixels in the vertical direction of the current pixel; And assigning distance weights to the edge interpolation value and the vertical interpolation value according to the distance between the current pixel and the pixel in the edge direction and the distance between the current pixel and the pixel in the vertical direction. The final interpolation calculation unit may be calculated.

Wherein the edge direction detection unit detects the edge direction according to an EDI method; The edge interpolation calculation unit and the vertical interpolation calculation unit may calculate the edge interpolation value and the vertical interpolation value, respectively, according to a NAL interpolation method.

Further, the edge interpolation calculation unit calculates the edge interpolation value according to the following equation: Xe (i, j) = (temp_e1 × 0.5 + temp_e2 × 0.5) × β + (Xe (j-1) × 0.5 + Xe (j + 1) ) × 0.5) (1-β), temp_e1 = Xe (j-1) × 1.5-Xe (j-3) × 0.5, temp_e2 = Xe (j + 1) × 1.5-Xe (j + 3) × 0.5, Where Xe (i, j) is the edge interpolation value of the current pixel, Xe (j-3), Xe (j-1), Xe (j + 1), Xe (j + 3) is the Value, β means weight) The vertical interpolation calculation unit calculates the vertical interpolation value according to the following equation: Xv (i, j) = (temp_v1 × 0.5 + temp_v2 × 0.5) × β + X (i, j-1) × 0.5 + X (i, j + 1) × 0.5) (1-β), temp_v1 = X (i, j-1) × 1.5-X (i, j-3) × 0.5, temp_v2 = X (i, j +1) × 1.5-X (i, j + 3) × 0.5, where Xv (i, j) is the vertical interpolation value of the current pixel, X (i, j-3), X (i, j-1 ), X (i, j + 1), X (i, j + 3) denote values of pixels in the vertical direction, and β denote weights.) The final interpolation calculation unit is based on the edge interpolation value and the vertical interpolation value. The value of the current pixel can be calculated according to the following equation: X (i, j) = (Xe (i, j) × d1 + Xv (i, j) × d2) / (d1 + d2), where d1 and d2 represent distance weights

Meanwhile, according to the present invention, there is provided an image interpolation method, comprising: detecting an edge direction of a current pixel; And assigning distance weights to the pixels in the edge direction and the pixels in the vertical direction according to the distance between the current pixel and the pixels in the edge direction and the distance between the current pixels and the pixels in the vertical direction. It can be achieved by an image interpolation method comprising the step of calculating the.

The interpolation of the current pixel may include calculating an edge interpolation value in consideration of a regional pattern based on the pixels in the edge direction; Calculating a vertical interpolation value in consideration of a regional pattern based on the pixels in the vertical direction; And assigning distance weights to the edge interpolation value and the vertical interpolation value according to the distance between the current pixel and the pixel in the edge direction and the distance between the current pixel and the pixel in the vertical direction. It may include the step of calculating.

The edge interpolation value is calculated according to the following equation: Xe (i, j) = (temp_e1 × 0.5 + temp_e2 × 0.5) × β + (Xe (j-1) × 0.5 + Xe (j + 1) × 0.5) 1-β), temp_e1 = Xe (j-1) × 1.5-Xe (j-3) × 0.5, temp_e2 = Xe (j + 1) × 1.5-Xe (j + 3) × 0.5, where Xe ( i, j is the edge interpolation value of the current pixel, Xe (j-3), Xe (j-1), Xe (j + 1), Xe (j + 3) is the value of pixels in the edge direction, β is the weight The vertical interpolation value is calculated according to the following equation: Xv (i, j) = (temp_v1 × 0.5 + temp_v2 × 0.5) × β + X (i, j-1) × 0.5 + X (i, j + 1) X0.5) (1-β), temp_v1 = X (i, j-1) × 1.5-X (i, j-3) × 0.5, temp_v2 = X (i, j + 1) × 1.5-X (i, j + 3) × 0.5, where Xv (i, j) is the vertical interpolation value of the current pixel, X (i, j-3), X (i, j-1), X (i, j + 1) X (i, j + 3) is the value of the pixels in the vertical direction, β denotes the weight. The value of the current pixel may be calculated as follows: X (i, j) = (Xe ( i, j) × d1 + Xv (i, j) × d2) / (d1 + d2), where d1 and d2 represent distance weights

On the other hand, according to the present invention, the object can be achieved by a computer-readable recording medium on which the method according to any one of the above is recorded.

As described above, according to the present invention, there is provided an image interpolation apparatus adopting a new distance weighting concept based on an interpolation scheme considering a regional pattern, a method thereof, and a computer-readable recording medium on which the method is recorded.

Hereinafter, specific embodiments of the present invention will be described with reference to the drawings.

1 is a schematic control block diagram of an image interpolation apparatus according to an embodiment of the present invention. As shown in FIG. 1, the image interpolation apparatus according to the present exemplary embodiment includes an edge direction detection unit 10 and a pixel interpolation unit 20.

The edge direction detection unit 10 is used to detect the direction of the edge of the current pixel. For example, an edge detection method used in edge-based line averaging (ELA), edge dependent interpolation (EDI) algorithm, or the like may be used.

The ELA algorithm detects the correlations of three cases in the vertical direction, the positive diagonal direction, and the negative diagonal direction using a 3x3 window, and finds the direction based on the correlation between adjacent pixels. In contrast, the EDI algorithm finds an edge direction using two vectors, which are sets of three pixels.

In general, pixel-by-pixel approaches are likely to lead to poor results due to noise, contrast variations, or inconspicuous edges. Therefore, in the present embodiment, an example of using the EDI method is to increase the probability of finding a reliable edge.

First, two vectors can be defined as

Here, X (i, j) represents the current pixel, i represents the position in the x-axis y-axis direction. Also, l and m are the positions of the vectors in the horizontal direction.

The edge direction detection unit 10 calculates a difference value between them based on these vectors according to the following equation.

Here, w is a weight according to the distance.

The edge direction detection unit 10 finally calculates (l ', m') according to the following equation.

2 illustrates the edge detection result of the edge direction detection unit 10. Referring to Fig. 2, the edge directions all have five cases.

When calculated as (l ', m') = (-1, -1), (0,0), (1,1), the edge direction is determined in the vertical direction (arrow direction indicated by vertical in FIG. 2). On the other hand, if (l ', m') = (-1,1) or (1, -1) is calculated, the edge direction is ± 45 degrees (arrow direction indicated by dia_10 or dia_20 in FIG. 2), (l ', If m ') = (-1,0), (0,1), or (0, -1), (1,0), the edge direction is ± 77.5 degrees (dia_11 and dia_12 or dia_21 and dia_21 in Figure 2). Arrow direction).

This is summarized as follows.

 The pixel interpolator 20 calculates a current pixel value based on neighboring pixel values using the detected edge direction information. The pixel interpolator 20 interpolates the current pixel by introducing a concept of distance weight. That is, in addition to the pixels in the edge direction, the pixels in the vertical direction closer together than the pixels in the edge direction are considered together, and at this time, the pixels in the edge direction and the pixels in the vertical direction are weighted according to the distance and presently. Calculate the pixel value.

As illustrated in FIG. 1, the pixel interpolation unit 20 according to the present exemplary embodiment includes an edge interpolation calculation unit 21, a vertical interpolation calculation unit 23, and a final interpolation calculation unit 25.

The edge interpolation calculation unit 21 calculates an edge interpolation value containing information on the edge direction based on the pixels in the edge direction according to the detected edge direction. The interpolation technique considering the regional pattern is used for the edge interpolation calculation unit 21 of the present invention. In the present embodiment, an edge interpolation value is calculated using a new adaptive linear interpolation (NAL) algorithm as an example.

3 shows the basic concept of the NAL algorithm.

Xe (i, j) represents the edge interpolation value of the current pixel, and Xe (j-3), Xe (j-1), Xe (j + 1), and Xe (j + 3) represent the reference pixel in the edge direction. Indicates the value of

For example, if the edge direction is detected at 45 °, X (i + 3, j-3), X (i + 1, j-1), X (i located in the arrow direction indicated by dia_20 in FIG. -1, j + 1) and X (i-3, j + 3) correspond to Xe (j-3), Xe (j-1), Xe (j + 1) and Xe (j + 3), respectively. .

The NAL algorithm uses four adjacent pixels, unlike the linear interpolation method using only two adjacent pixels.

The following equation is for calculating the edge interpolation value used by the edge interpolation calculation unit 21.

Xe (i, j) = (temp_e1 × 0.5 + temp_e2 × 0.5) × β + (Xe (j-1) × 0.5 + Xe (j + 1) × 0.5) (1-β)

Here, temp_e1 and temp_e2 are determined by the following equation.

temp_e1 = Xe (j-1) × 1.5-Xe (j-3) × 0.5

temp_e2 = Xe (j + 1) × 1.5-Xe (j + 3) × 0.5

In Equation 4, the beta value has a value between 0 and 1 as an adjustment factor. In the present invention, the beta value can be adjusted appropriately to obtain the best performance.

As can be seen from the above equation, the edge interpolation calculation unit 21 calculates temp_e1 and temp_e2 by giving weights to two pixels close to the current pixel among the pixels in four edge directions (1.5 in the above equation). The edge interpolation values are calculated by assigning weights β and 1-β to the average and linear interpolation values.

The above interpolation value calculation method reflects a regional pattern, and since the slope of the straight line in FIG. 3 may be 0 ° or 90 °, in order to solve this problem, a weight (β) is adjusted by using a linear interpolation value to adjust it. do.

Meanwhile, the vertical interpolation calculator 23 calculates a vertical interpolation value based on the pixels in the vertical direction of the current pixel. The vertical interpolation calculation unit 23 of the present invention calculates the vertical interpolation value using the NAL algorithm in the same manner as the above-described calculation method of the edge interpolation calculation unit 21.

This is expressed as follows.

Xv (i, j) = (temp_v1 × 0.5 + temp_v2 × 0.5) × β + X (i, j-1) × 0.5 + X (i, j + 1) × 0.5) (1-β)

Here, temp_v1 and temp_v2 are determined by the following equation.

temp_v1 = X (i, j-1) × 1.5-X (i, j-3) × 0.5

temp_v2 = X (i, j + 1) × 1.5-X (i, j + 3) × 0.5

Where Xv (i, j) represents a vertical interpolation value, and X (i, j-3), X (i, j-1), X (i, j + 1) and X (i, j + 3) Denotes values of reference pixels in the vertical direction of FIG. 2. Since the above equation is similar to the above Equations 4 and 5, a detailed description thereof will be omitted.

The final interpolation calculator 25 provides distance weights to the edge interpolation values and the vertical interpolation values calculated above, and finally calculates the current pixel values. As described above, unlike the conventional method of interpolating the current pixel by considering only the pixels in the edge direction, by introducing the concept of distance weight, the values of the pixels in the vertical direction near to the current pixel are considered more accurate. Enables interpolation of the current pixel to values.

Here, the distance weight is given according to the distance between the current pixel, and the value is given according to the distance between the current pixel and the pixel in the edge direction and the distance between the current pixel and the pixel in the vertical direction.

The final interpolation value calculation method of the final interpolation calculation unit 25 may be expressed by the following equation.

X (i, j) = (Xe (i, j) × d1 + Xv (i, j) × d2) / (d1 + d2)

Here, d1 and d2 mean distance weights.

In this embodiment, since there are five edge directions in total, a method of interpolating the current pixel values for each of them will be described.

First example

As a first example, the edge direction is detected in the vertical direction.

In this case, the edge interpolation value calculated according to Equations 4 and 5 and the vertical interpolation value calculated according to Equations 6 and 7 are calculated to be the same value. In addition, since the distances are the same, the distance weights are given the same, resulting in the same value.

That is, X (i, j) = Xe (i, j) = Xv (i, j) is determined.

Second example

As a second example, the edge direction is detected at 45 degrees.

In this case, the pixels in the edge direction are X (i + 3, j-3), X (i + 1, j-1), X (i-1, j + 1), located in the dia_20 arrow direction in FIG. 2. X (i-3, j + 3).

Thus, temp_e1 and temp_e2 are respectively calculated as follows:

temp_e1 = X (i + 1, j-1) × 1.5-X (i + 3, j-3) × 0.5,

temp_e2 = Xe (i-1, j + 1) × 1.5-X (i-3, j + 3) × 0.5

Accordingly, the edge interpolation value is calculated as follows:

Xe (i, j) = (temp_e1 × 0.5 + temp_e2 × 0.5) × β + (X (i + 1, j-1) × 0.5 + Xe (i + 1, j + 1) × 0.5) (1-β).

On the other hand, the vertical interpolation value is calculated in the same manner as in the first example according to equations (6) and (7).

, d2=1로 결정된다. 이는 현재 화소와 에지방향의 화소들과의 최소거리가

이고, 현재 화소와 수직방향의 화소들과의 최소거리가 1이기 때문이다(도 4의 (a) 참조).In this example, the distance weight d1 =

, d2 = 1. This means that the minimum distance between the current pixel and the pixels in the edge direction

This is because the minimum distance between the current pixel and the pixels in the vertical direction is 1 (see FIG. 4A).

Thus, the final current pixel value is calculated as follows:

Third example

As a third example, the edge direction is detected at -45 degrees.

In this case, the pixels in the edge direction are X (i-3, j-3), X (i-1, j-1), X (i + 1, j + 1), located in the dia_10 arrow direction in FIG. X (i + 3, j + 3) is obtained.

Thus, temp_e1 and temp_e2 are respectively calculated as follows:

temp_e1 = X (i-1, j-1) × 1.5-X (i-3, j-3) × 0.5,

temp_e2 = Xe (i + 1, j + 1) × 1.5-X (i + 3, j + 3) × 0.5

Accordingly, the edge interpolation value is calculated as follows:

Xe (i, j) = (temp_e1 × 0.5 + temp_e2 × 0.5) × β + (X (i + 1, j-1) × 0.5 + Xe (i + 1, j + 1) × 0.5) (1-β)

On the other hand, the vertical interpolation value is calculated in the same manner as in the first example according to equations (6) and (7).

, d2=1 로 결정된다. 따라서, 최종적인 현재 화소의 값은 다음과 같이 산출된다:In this example, the distance weight is determined in the same manner as in the second example. That is, d1 =

, d2 = 1. Thus, the final value of the current pixel is calculated as follows:

Fourth example

As a fourth example, the edge direction is detected at 77.5 degrees.

In this case, the pixels in the edge direction are X (i + 1, j-3), X (i, j-1), X (i-1, j + 1), and X ( i-2, j + 3) and X ((i + 2, j-3), X (i + 1, j-1), X (i, j + 1) located in the dia_21 arrow direction in FIG. In this case, the final interpolation value is calculated by averaging the interpolation values for each of them.

That is, edge interpolation values (Xe1 based on X (i + 1, j-3), X (i, j-1), X (i-1, j + 1), and X (i-2, j + 3) (i, j)) and the final interpolation value 1 (X1 (i, j)) according to the vertical interpolation value Xv (i, j), and X (i + 2, j-3), X (i + 1 edge interpolation value Xe2 (i, j) and vertical interpolation value Xv (i, j) based on, j-1), X (i, j + 1), X (i-1, j + 3) The average value of the final interpolation value 2 (X2 (i, j)) is determined as the value of the current pixel.

, d2=1로 결정한다. 비록 현재 화소와 에지방향의 화소의 최소거리는 1이지만, 다른 예들하고는 달리 에지방향의 화소들 모두가 현재 화소를 중심으로 대칭적으로 위치하지 않기 때문에(즉, 현재 화소를 지나는 에지방향에 위치하지 않음), 현재 화소를 지나는 에지방향으로 화소가 있다고 가정하고 그 거리를 산출한다. 도 4의 (b) 를 참조하면 수직으로는 1, 수평으로는 0.5이므로, 가상의 에지방향의 화소까지의 거리는

가 되므로, 이를 거리 가중치로 결정한다.Here, when calculating the final interpolation value 1 and the final interpolation value 2, the distance weight value d1 =

, d2 = 1. Although the minimum distance between the current pixel and the pixel in the edge direction is 1, unlike the other examples, since all of the pixels in the edge direction are not symmetrically positioned about the current pixel (that is, they are not positioned in the edge direction past the current pixel). And assuming that there is a pixel in the edge direction passing through the current pixel, the distance is calculated. Referring to FIG. 4B, since it is 1 vertically and 0.5 horizontally, the distance to the pixel in the virtual edge direction is

Since this is determined as the distance weight.

The above expression is expressed as:

5th example

As a last example, the edge direction is detected as -77.5 °. In this case, the pixels in the edge direction are X (i-2, j-3), X (i-1, j-1), X (i, j + 1), and X ( i + 1, j + 3) and X (i-1, j-3), X (i, j-1), X (i + 1, j + 1) and X (i + located in dia_12 direction 2, j + 3) There are two things. Therefore, similarly to the fourth example, the final interpolation value is calculated as an average of the interpolation values for each of these.

, d2=1로 결정한다. At this time, the distance weight is the same as the fourth example, d1 =

, d2 = 1.

The final interpolation value can be obtained by equation (8) in the fourth example. Since the calculation method is similar to the fourth example, a detailed description thereof will be omitted.

 Hereinafter, an experimental result of applying an image interpolation method according to an embodiment of the present invention to actual image processing will be described.

In this experiment, we compared the conventional image interpolation algorithm in terms of objective performance and subjective image quality of the image interpolation method according to the present invention. The experiments included eight still images, including the image named 'Barbara', and first down-sampled the original image of 512 × 512 size. That is, only the even lines of the original image are copied to form a 512 × 256 size image, and then the present invention is applied to interpolating odd lines.

However, since the interpolation method of the present invention applies the NAL algorithm, the performance may vary depending on the β value. Also, β value representing optimal performance may vary depending on the image. The table below shows β values showing optimal results in each test image, and FIG. 5 is a graph thereof.

As can be seen from Table 2 and Figure 5, it can be seen that the objective performance (PSNR) is steadily improved without a significant difference in the change of the adjustment factor β value. In addition, the average value of β shows an optimal result. Therefore, β value was applied to 0.3 in this experiment.

The table below shows the results of image processing using the conventional interpolation method and the image interpolation method of the present invention in still images.

As can be seen from the above table, it can be seen that the image interpolation method of the present invention is superior in objective performance compared to the conventional image interpolation method. The PSNR value occupies the 1st and 2nd in every image of the subject.

6A to 6G show images of processing results for comparing subjective image quality between the present invention and the conventional method.

6A is an original image, and FIG. 6B is an enlarged image of a partial region of FIG. 6A. FIG. 6C is a result image of applying the line doubling (LD) method, and FIG. 6D is a result image of applying the line averaging (LA) method. FIG. 6E shows an ELA method, FIG. 6F shows an E-ELA method, FIG. 6G shows an EDI method, FIG. 6H shows a DOI method, and FIG. 6I shows an image obtained by applying the image interpolation method of the present invention.

6A to 6I, it can be seen that the image interpolation method according to the present invention is superior to the DOI algorithm which is evaluated to have a relatively high computational complexity and excellent performance in terms of subjective image quality. In addition, it can be seen that the image to which the present invention is applied is almost similar to the original image.

The table below compares the execution time for image processing by each interpolation method to measure the complexity of calculations through various interlaced scanning techniques (Barbara images were used).

As can be seen from the above table, the image interpolation method according to the present invention takes some time compared to some conventional algorithms, but has a processing speed about five times faster than the DOI algorithm which is evaluated as excellent. As described above, since the image interpolation method according to the present invention is superior in performance to the conventional interpolation algorithm, it may be understood that it takes some processing time.

The image interpolation method according to an embodiment of the present invention described above may be implemented as a computer readable program. In addition, the components included in the image interpolation apparatus according to the embodiment of the present invention described above may be manufactured by a computer readable program, and the image interpolation apparatus may include a storage device such as a memory in which the program is installed, and the program. It can be implemented by a computer device including a running processor and the like. In addition, it is applicable to all displays equipped with a processor that receives and interpolates interlaced signals.

In the above-described embodiment, the value of the distance weight is described as being directly proportional to the distance between the current pixel and the corresponding pixels. However, the distance weight value may not be given in proportion to the distance, but may be given approximately or differently. .

Although some embodiments of the invention have been shown and described, it will be apparent to those skilled in the art that modifications may be made to the embodiment without departing from the spirit or spirit of the invention. . It is intended that the scope of the invention be defined by the claims appended hereto and their equivalents.

1 is a schematic control block diagram of an image interpolation apparatus according to an embodiment of the present invention;

2 is a schematic diagram for explaining an edge direction detected by an image interpolation apparatus according to an embodiment of the present invention;

3 is a schematic diagram for explaining an NAL algorithm used in an image interpolation apparatus according to an embodiment of the present invention;

4 shows an example of calculating distance weights according to an embodiment of the present invention;

5 is a graph comparing performance according to adjustment factors in an image interpolation apparatus according to an embodiment of the present invention;

6A to 6I illustrate comparative images of an image to which the present invention is applied and an image to which a conventional interpolation method is applied.

Claims (8)

In the video interpolation device, An edge direction detection unit detecting an edge direction of the current pixel; And A pixel beam that calculates a value of the current pixel by assigning distance weights according to a distance between the current pixel and a pixel adjacent to the current pixel in the edge direction and a distance between the current pixel and a pixel adjacent to the current pixel in a vertical direction An image interpolation apparatus comprising an executive. The method of claim 1, The pixel interpolation unit may include an edge interpolation calculation unit configured to calculate an edge interpolation value based on pixels positioned in the edge direction; A vertical interpolation calculator configured to calculate a vertical interpolation value based on pixels positioned in the vertical direction of the current pixel; And A distance weight is assigned to the edge interpolation value and the vertical interpolation value according to a distance between the current pixel and a pixel adjacent to the current pixel in the edge direction, and a distance between the current pixel and a pixel adjacent to the current pixel in the vertical direction. And a final interpolation calculator configured to calculate a value of the current pixel. The method of claim 2, The edge direction detection section detects the edge direction according to an EDI method; And the edge interpolation calculation unit and the vertical interpolation calculation unit calculate the edge interpolation value and the vertical interpolation value, respectively, according to a NAL interpolation method. The method of claim 3, The edge interpolation calculation unit calculates the edge interpolation value according to the following equation: Xe (i, j) = (temp_e1 × 0.5 + temp_e2 × 0.5) × β + (Xe (j-1) × 0.5 + Xe (j + 1) × 0.5) (1-β), temp_e1 = Xe (j-1) × 1.5-Xe (j-3) × 0.5, temp_e2 = Xe (j + 1) × 1.5-Xe (j + 3) × 0.5, Where Xe (i, j) is the edge interpolation value of the current pixel, Xe (j-3), Xe (j-1), Xe (j + 1), and Xe (j + 3) are located in the edge direction. Value of pixels, β means weight) The vertical interpolation calculation unit calculates the vertical interpolation value according to the following equation: Xv (i, j) = (temp_v1 × 0.5 + temp_v2 × 0.5) × β + X (i, j-1) × 0.5 + X (i, j + 1) × 0.5) (1-β), temp_v1 = X (i, j-1) × 1.5-X (i, j-3) × 0.5, temp_v2 = X (i, j + 1) × 1.5-X (i, j + 3) × 0.5, Where Xv (i, j) is the vertical interpolation of the current pixel, X (i, j-3), X (i, j-1), X (i, j + 1), X (i, j + 3) is the value of the pixels located in the vertical direction, β means the weight) And the final interpolation calculator calculates a value of the current pixel based on the edge interpolation value and the vertical interpolation value according to the following equation: X (i, j) = (Xe (i, j) × d1 + Xv (i, j) × d2) / (d1 + d2) Where d1 and d2 represent distance weights In the image interpolation method, Detecting an edge direction of a current pixel; And Calculating a value of the current pixel by assigning a distance weight to the distance between the current pixel and a pixel adjacent to the current pixel in the edge direction and a distance between the current pixel and a pixel adjacent to the current pixel in a vertical direction. Image interpolation method comprising a. The method of claim 5, The interpolation of the current pixel may include calculating an edge interpolation value considering a local pattern based on pixels located in the edge direction; Calculating a vertical interpolation value in consideration of a regional pattern based on the pixels located in the vertical direction; And A distance weight is assigned to the edge interpolation value and the vertical interpolation value according to a distance between the current pixel and a pixel adjacent to the current pixel in the edge direction, and a distance between the current pixel and a pixel adjacent to the current pixel in the vertical direction. And calculating a value of the current pixel. The method of claim 6, The edge interpolation value is calculated according to the following equation: Xe (i, j) = (temp_e1 × 0.5 + temp_e2 × 0.5) × β + (Xe (j-1) × 0.5 + Xe (j + 1) × 0.5) (1-β), temp_e1 = Xe (j-1) × 1.5-Xe (j-3) × 0.5, temp_e2 = Xe (j + 1) × 1.5-Xe (j + 3) × 0.5, Where Xe (i, j) is the edge interpolation value of the current pixel, Xe (j-3), Xe (j-1), Xe (j + 1), and Xe (j + 3) are located in the edge direction. Value of pixels, β means weight) The vertical interpolation value is calculated according to the following equation: Xv (i, j) = (temp_v1 × 0.5 + temp_v2 × 0.5) × β + X (i, j-1) × 0.5 + X (i, j + 1) × 0.5) (1-β), temp_v1 = X (i, j-1) × 1.5-X (i, j-3) × 0.5, temp_v2 = X (i, j + 1) × 1.5-X (i, j + 3) × 0.5, Where Xv (i, j) is the vertical interpolation of the current pixel, X (i, j-3), X (i, j-1), X (i, j + 1), X (i, j + 3) is the value of the pixels located in the vertical direction, β means the weight) An image interpolation method, wherein the value of the current pixel is calculated as follows: X (i, j) = (Xe (i, j) × d1 + Xv (i, j) × d2) / (d1 + d2) Where d1 and d2 represent distance weights A computer-readable recording medium having recorded thereon a method according to any one of claims 5 to 7.

Images