KR101493425B1 - Apparatus and method for interpolating image, and apparatus for processing image using the same - Google Patents

Abstract

The present invention relates to an image interpolation apparatus, an image interpolation method, and an image processing apparatus using the same. According to an aspect of the present invention, there is provided an image interpolation apparatus including: an edge region identification unit for identifying an edge region including an edge from an image; An edge region classifying unit classifying the edge region into a plurality of regions; And an interpolation section for performing interpolation in different ways for each of the zones.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image interpolation apparatus, an image interpolation method, and an image processing apparatus using the interpolation apparatus,

The present invention relates to an image interpolation apparatus, an image interpolation method, and an image processing apparatus using the same.

In the image processing, enlarging the original image means shifting the coordinates of each pixel of the original image by a magnification for enlarging the original image. In this process, pixels that have not been assigned pixel values exist in the enlarged image. These pixels are called holes. The technique of assigning appropriate values to empty pixels and processing images is called interpolation.

Interpolation is performed by a linear function or a polynomial regression analysis, Fourier series, spline, moving average, Kriging, etc., using values of known pixels. The interpolation method includes Nearest Neighbor Interpolation, Bilinear Interpolation, Cubic Interpolation, Higher Order Interpolation, and B-Spline Interpolation. have.

However, the interpolated image using the above-described method has poor image quality as compared with the original image. As the enlargement ratio of the image increases, the detail of the image becomes difficult to understand. Particularly, blocking occurs at a portion corresponding to an edge of the image, and it is difficult to expect a high-quality image according to the conventional interpolation method.

An object of the present invention is to provide an image interpolating apparatus, an image interpolating method, and an image processing apparatus for obtaining an image in which image quality is not greatly degraded as compared with an original image.

It is another object of the present invention to provide an image interpolating apparatus, an image interpolating method, and an image processing apparatus that allow an edge portion of an image to be clearly restored even after interpolation.

Another object of the present invention is to provide an image interpolating apparatus, an image interpolating method, and an image processing apparatus which do not greatly increase the amount of calculation required for interpolation while obtaining a high-quality image.

According to an aspect of the present invention, there is provided an image interpolation apparatus including: an edge region identification unit for identifying an edge region including an edge from an image; An edge region classifying unit classifying the edge region into a plurality of regions; And an interpolation section for performing interpolation in different ways for each of the zones.

The edge region identification unit may identify the edge region by calculating a variance of pixel values of pixels included in one region of the image.

Wherein the edge area identification unit determines the one area as an edge area if the variance of the pixel value is greater than a preset threshold value and if the variance of the pixel value is less than or equal to the threshold value, You can decide.

The image interpolating apparatus may further include an edge map generating unit for detecting an edge included in the edge area and generating an edge map.

Wherein the edge map generation unit comprises: an edge detection unit configured to detect a first pixel corresponding to an edge in the edge region, detect a second pixel corresponding to an edge of a plurality of pixels located around the first pixel, The second pixel can be connected.

The edge region classifying unit may classify the edge regions according to the number of edges included in one region of the edge map.

Wherein the edge region classifier determines the first zone as a first zone if the number of edges included in the one zone is less than or equal to a preset threshold value and if the number of edges included in the one zone is greater than the threshold value, The one zone may be determined as the second zone.

The interpolator may use more than four adjacent pixels located around the pixel to interpolate pixels contained in the first region.

The interpolator may interpolate the pixel using twelve adjacent pixels surrounding the pixels included in the first zone.

The interpolator may calculate a pixel value of the pixel by applying a weight to pixel values of each of the plurality of adjacent pixels, and summing the weight values.

The weight can be calculated using a Wiener filter.

The weight may be calculated based on a distance between the pixel and the adjacent pixel.

The interpolator may use less than four adjacent pixels located around the pixel to interpolate the pixels contained in the second region.

The interpolator may use a linear interpolation scheme to interpolate pixels in the non-edge region that do not include an edge in the image.

The image interpolation method according to an embodiment of the present invention can perform interpolation in a plurality of ways for an area including an edge in an image using an image interpolator.

The image interpolation method comprising: identifying an edge region from the image; Classifying the edge region into a plurality of regions; And performing interpolation in different ways for each of the zones.

Identifying the edge region comprises: retrieving an area of predetermined size from the image; Calculating a variance of pixel values of pixels included in the loaded area; And determining the referred area as an edge area if the variance of the pixel value is greater than a predetermined threshold value.

The step of classifying the edge region into a plurality of regions includes: generating an edge map by detecting an edge included in the edge region; Loading a zone of predetermined size from the edge map; Calculating the number of edges included in the loaded area; Determining the loaded zone as a first zone if the number of edges is less than or equal to a predetermined threshold; And determining the loaded area as the second zone if the number of edges is greater than a predetermined threshold value.

Performing the interpolation may comprise: interpolating the pixel using more than four adjacent pixels located around the pixels included in the first zone.

Performing the interpolation may comprise: interpolating the pixel using less than four adjacent pixels located around the pixels included in the second region.

Performing the interpolation may comprise: interpolating the pixels of the non-edge region that do not include an edge in the image in a linear interpolation manner.

An image processing apparatus according to an embodiment of the present invention includes: a receiving unit for receiving an image; A processing unit for processing the received image; A storage unit for storing data used for image or image processing; And an output unit outputting the processed image, wherein the processing unit comprises: an edge area identification unit for identifying an edge area including an edge from the received image; An edge map generation unit for generating an edge map by detecting an edge included in the edge region; An edge region classifying unit classifying the edge region into a plurality of regions using the edge map; And an interpolation section for performing interpolation in different ways for each of the zones.

The edge region classifying section may determine that the first zone is a first zone if the number of edges included in one zone of the edge zone is less than or equal to a preset threshold value and if the number of edges included in one zone of the edge zone is If it is larger than the threshold value, the zone can be determined as the second zone.

Wherein the interpolator performs interpolation using more than four adjacent pixels located around the pixel for the pixels included in the first zone and for interpolating pixels included in the second zone around the pixel An interpolation can be performed using a linear interpolation method for pixels in a non-edge area that does not include edges in the received image.

The image interpolation method according to an embodiment of the present invention may be implemented as a program that can be executed in a computer and recorded in a computer-readable recording medium.

According to an embodiment of the present invention, even after the interpolation, an image in which the image quality is not significantly degraded as compared with the original image can be obtained.

Also, according to an embodiment of the present invention, an edge portion of an image can be vividly reconstructed even after interpolation.

According to an embodiment of the present invention, a high-quality image can be obtained without greatly increasing the amount of calculation required for interpolation.

1 is a block diagram illustrating an image interpolation apparatus according to an embodiment of the present invention.
2 shows an example of an image to be interpolated according to an embodiment of the present invention.
FIG. 3 illustrates an example of an image in which an edge region and a non-edge region are separated according to an embodiment of the present invention.
FIG. 4 illustrates an example of an image in which an edge region is divided into a plurality of regions according to an embodiment of the present invention.
5 is a diagram illustrating a process of generating an edge map according to an embodiment of the present invention.
FIGS. 6 and 7 illustrate a process of classifying edge regions according to an embodiment of the present invention.
8 is a diagram illustrating a process of interpolating pixels included in an edge region according to an embodiment of the present invention.
9 to 13 are diagrams illustrating a process of interpolating pixels included in an edge region according to another embodiment of the present invention.
FIG. 14 is a flowchart illustrating an image interpolation method according to an embodiment of the present invention.
15 is a flowchart illustrating steps for identifying an edge region from an image in accordance with an embodiment of the present invention.
16 is a flow chart illustrating the step of classifying an edge region into a plurality of regions according to an embodiment of the present invention.
17 is a block diagram showing an image processing apparatus according to an embodiment of the present invention.

Other advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail with reference to the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the invention to those skilled in the art. Is provided to fully convey the scope of the invention to those skilled in the art, and the invention is only defined by the scope of the claims.

Unless defined otherwise, all terms (including technical or scientific terms) used herein have the same meaning as commonly accepted by the generic art in the prior art to which this invention belongs. Terms defined by generic dictionaries may be interpreted to have the same meaning as in the related art and / or in the text of this application, and may be conceptualized or overly formalized, even if not expressly defined herein I will not.

The terminology used herein is for the purpose of illustrating embodiments and is not intended to be limiting of the present invention. In the present specification, the singular form includes plural forms unless otherwise specified in the specification. As used herein, the terms' comprise 'and / or various forms of use of the verb include, for example,' including, '' including, '' including, '' including, Steps, operations, and / or elements do not preclude the presence or addition of one or more other compositions, components, components, steps, operations, and / or components. The term 'and / or' as used herein refers to each of the listed configurations or various combinations thereof.

It should be noted that the terms such as '~', '~ period', '~ block', 'module', etc. used in the entire specification may mean a unit for processing at least one function or operation. For example, a hardware component, such as a software, FPGA, or ASIC. However, '~ part', '~ period', '~ block', '~ module' are not meant to be limited to software or hardware. Modules may be configured to be addressable storage media and may be configured to play one or more processors. ≪ RTI ID = 0.0 >

Thus, by way of example, the terms 'to', 'to', 'to block', 'to module' refer to components such as software components, object oriented software components, class components and task components Microcode, circuitry, data, databases, data structures, tables, arrays, and the like, as well as components, Variables. The functions provided in the components and in the sections ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ' , '~', '~', '~', '~', And '~' modules with additional components.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings attached hereto.

An image interpolation apparatus according to an exemplary embodiment of the present invention may perform interpolation in a plurality of ways with respect to an edge region including an edge in an image. In other words, the image interpolation apparatus according to an embodiment of the present invention can classify the edge region of the image into a plurality of regions, and selectively apply an interpolation method to each region.

1 is a block diagram illustrating an image interpolation apparatus according to an embodiment of the present invention. 1, an image interpolation apparatus 100 according to an embodiment of the present invention may include an edge region identification unit 11, an edge region classification unit 13, and an interpolation unit 14. The edge region identification section 11 can identify an edge region including an edge from the image. The edge region classifying unit 13 may classify the edge region into a plurality of regions. The interpolator 14 may perform interpolation in different ways for each of the zones.

The edge region identification unit 11 can identify and distinguish an edge region including an edge and a non-edge region including no edge from an image to be interpolated. 2 shows an example of an image to be interpolated according to an embodiment of the present invention.

Referring to FIG. 2, the image 20 shows a house 22 on a meadow 21. The grasses (21) are dense with grasses, and the image (20) details the appearance of the grasses (21) where the grasses grow.

According to an embodiment of the present invention, the edge region identification unit 11 may identify an edge region including an edge from the image 20. [ For example, as shown in FIG. 3, the edge region identification section 11 can identify an edge region 201 that includes an edge in the image 20. FIG. In the image 20, the remaining area excluding the edge area 201 is a non-edge area 202, 203, 204, or 205 that does not include an edge. In other words, in the image (20), sky, house roof, walls and doors are expressed in color only, without edges.

According to an embodiment of the present invention, the edge region identification unit 11 may identify the edge region 201 by calculating variances of pixel values of pixels included in one region of the image 20 . According to an exemplary embodiment, the edge region identification unit 11 may determine a region included in one region of the image 20 as an edge region if the variance of pixel values is greater than a preset threshold value. In contrast, if the variance of pixel values of pixels included in one area of the image 20 is less than or equal to the threshold value, the edge area identification unit 11 may determine the area as a non-edge area.

One region of the image for which the variance is calculated to identify the edge region 201 may be 4 x 4 in size, but is not so limited and the variance can be calculated for regions of smaller or larger size.

If the variance of pixel values in a given region is larger than a preset threshold value, it may be considered that the difference in pixel value between the pixels is large, so that the region may be classified into the edge region 201 including the edge. Conversely, if the variance of pixel values in a given region is below a threshold, it can be assumed that the pixel values of the pixels are evenly distributed over the region, so that region can be classified as a non-edge region that does not include an edge.

The edge region classifying unit 13 may classify the edge region 201 into a plurality of regions. FIG. 4 shows an example of an image 20 in which the edge region 201 is divided into two regions 201a and 201b according to an embodiment of the present invention. 4, the edge region 201 includes a main zone 201a that includes a main edge that contributes to the main contour of the image, and a minor zone 201b that includes a minor edge that contributes to the detail of the image. . ≪ / RTI > As shown in FIG. 4, the edge region classification unit 13 may classify the edge region 201 into two regions, but the number of the regions to be classified is not limited thereto and may be three or more .

According to an embodiment of the present invention, the edge region classifying unit 13 may classify edge regions according to the number of edges included in one region of the edge region 201. According to one embodiment, when the number of edges included in one region of the edge region 201 is less than or equal to a predetermined threshold value, the edge region classification unit 13 divides the region into the first region, that is, the main region 201a). Conversely, if the number of edges included in one region of the edge region 201 is greater than the threshold value, the edge region classification unit 13 may determine the region as the second region, that is, the sub region 201b.

The more edges the edge region 201 contains in a region of constant size, the shorter the edge may be, which is likely to be a minor edge contributing to the detail of the image 20 . Also, the less edge the edge region 201 contains in a region of constant size, the longer edge may be the longer edge, which may be the main edge contributing to the main contour of the image 20 This is big. Thus, one embodiment of the present invention is to divide the edge region 201 into a plurality of regions, count the number of edges included in each region, and if the number of edges is less than the threshold, , And if the number of edges is larger than the threshold value, the zone can be classified as the sub-zone 201b.

1, the image interpolation apparatus 100 according to an embodiment of the present invention further includes an edge map generation unit 12 for detecting an edge included in the edge region 201 and generating an edge map . The edge map generation unit 12 assigns 1 to pixels corresponding to edges in the edge region 201 and 0 to pixels not corresponding to edges to indicate the distribution of edges in the edge region 201 The edge map can be generated.

5 is a diagram illustrating a process of generating an edge map according to an embodiment of the present invention. According to an embodiment of the present invention, the edge map generation unit 12 detects a first pixel corresponding to an edge in the edge area 201, and detects an edge of a plurality of pixels located around the first pixel , And may connect the first pixel and the second pixel.

For example, as shown in FIG. 5, the edge map generation unit 12 may detect a first pixel 31 corresponding to an edge in the edge region 201. The edge map generator 12 may then detect the second pixel corresponding to the edge of the eight pixels located around the first pixel 31 with the first pixel 31 as a starting point . Referring to FIG. 5, a pixel 32 located to the right of the first pixel 31 corresponds to an edge. The edge map generator 12 may connect the pixels corresponding to the edges by connecting the first pixel 31 corresponding to the starting point and the second pixel 32 positioned around the first pixel 31. [

Then, the edge map generating unit 12 may detect another pixel corresponding to the edge among the pixels located around the second pixel 32. [0050] Referring to FIG. 5, a pixel 33 located to the right of the second pixel 32 corresponds to an edge. The edge map generating unit 12 may connect the second pixel 32 with a third pixel 33 which is an edge pixel.

The edge map generating unit 12 may repeat the process described above until no new edge pixels are detected around the edge pixels. As a result, in the edge region 201 shown in Fig. 5, the detection of the edge 1 is completed after the fifth pixel 35 is completed.

Then, the edge map generation unit 12 detects a pixel corresponding to another edge in the edge region 201, detects an edge pixel located around the edge of the pixel as a starting point, and repeats the process of connecting them . For example, in the edge region 201 shown in Fig. 5, the edge map generation section 12 detects a first pixel 36 corresponding to another edge 2, and the first pixel 36 The second pixel 37 corresponding to the edge of the eight pixels positioned around the edge of the second pixel 37 is detected. The edge map generating unit 12 may connect the first pixels 36 and the second pixels 37 to connect the edge pixels. By repeating the above-described process, the edge map generating unit 12 can complete the detection of the edge 2 at the end of the fourth pixel 39. As described above, the edge map generation unit 12 detects pixels corresponding to the edges with respect to the edge region 201 and connects them to each other to generate an edge map indicating information on the distribution, length, and number of edges in the edge region Can be generated.

According to an embodiment of the present invention, the edge region classifying unit 13 may classify the edge region 201 into a plurality of regions using the edge map. According to one embodiment, if the number of edges included in one section of the edge map is less than or equal to a predetermined threshold value, the edge region classifying section 13 divides the zone into the first zone, that is, the main zone 201a, . Conversely, if the number of edges included in one section of the edge map is larger than the threshold value, the edge region classifying section 13 may determine the section as a second section, that is, a sub-section 201b.

The size of one zone where the number of edges is counted to classify the edge zone 201 may be 15 X 15, but the size of the zone is not limited thereto, and the number of edges may be counted for a zone of size smaller or larger .

FIGS. 6 and 7 are diagrams for explaining a process of classifying an edge region 201 into a plurality of regions according to an embodiment of the present invention. As shown in FIGS. 6 and 7, the edge region classifying section 13 retrieves a region of a predetermined size (for example, 15 X 15) from the edge map and counts the number of edges included in the region . The area shown in Fig. 6 includes a total of two edges 41, 42, and the area shown in Fig. 7 includes a total of six edges 43-48.

The edge area classifying unit 13 may count the number of edges included in the loaded area and compare the number of edges with a threshold value to classify the corresponding area. According to one embodiment, the threshold may be set to any one of 1 to 5, but is not limited thereto and may be set to be larger than 5 according to an image to be interpolated. In the embodiment shown in Figs. 6 and 7, it is assumed that the threshold value is set to 2.

Since the area shown in FIG. 6 includes two edges 41 and 42 of the same number as the threshold value, the edge area classification unit 13 can determine the area as the first area. Since the zone shown in Fig. 7 includes six edges (43 to 48) having a number larger than the threshold value, the edge zone classification unit 13 can determine the zone as the second zone.

According to another embodiment of the present invention, the edge region classifying unit 13 may classify the edge region based on the length of the edge, instead of the number of edges included in one region of the edge map. For example, according to another embodiment of the present invention, if the length of the longest edge among the edges included in one section of the edge map is longer than a preset threshold value, 1 zone 201a. If the edge length of the edge included in one segment of the edge map is shorter than or equal to the threshold value, the edge region classifying unit 13 may determine the zone as the second zone 201b.

According to another embodiment of the present invention, the edge region classifying section 13 may classify the edge region based on the entire length of edges included in one region of the edge map. For example, according to another embodiment of the present invention, the edge region classifying section 13 sums the lengths of edges included in one region of the edge map, and if the summed length is longer than the predetermined threshold value, May be determined as the first zone 201a, and if the summed length is shorter than or equal to the threshold value, the zone may be determined as the second zone 201b.

The interpolator 14 may perform interpolation in different ways for each of the areas classified by the edge area classifier 13. [ According to one embodiment of the present invention, the interpolating unit 14 uses more than four adjacent pixels located around the pixel to interpolate the pixels included in the first region 201a of the edge region 201 .

FIG. 8 is a diagram illustrating a process of interpolating pixels included in the first region 201a according to an embodiment of the present invention. 8, in accordance with an embodiment of the present invention, the interpolator 14 includes a plurality of pixels 50, which are arranged in a matrix, for interpolating the pixels 50 contained in the first zone 201a, Twelve adjacent pixels 501 to 512 may be used. In other words, the interpolator 14 may refer to the pixel values of the pixels located in the 12-direction around the pixel in order to interpolate the pixel 50 included in the first area 201a.

According to an embodiment of the present invention, the interpolator 14 may calculate the pixel values of the pixels 50 by applying weights to the pixel values of the adjacent pixels 501 to 512 and summing them. For example, the interpolator 14 may interpolate the pixel 50 using the following equation:

Where α is the weight applied to the pixel value, and α _{0, 3, 12, and 15} are zero.

According to an embodiment of the present invention, the weight a may be calculated using a Wiener filter theory. For example, the weight alpha can be calculated using the following equation: < RTI ID = 0.0 >

은 기설정된 사이즈를 갖는 영역에 포함된 픽셀들의 cross covariance 행렬을 나타낸다. 본 발명의 일 실시예에 따르면, 상기 가중치 α를 계산하기 위해 R 및

가 계산되는 영역의 사이즈는, 상기 에지 영역 분류부(13)에 의해 에지의 개수가 카운팅되는 구역의 사이즈와 동일할 수 있다.Here, R represents an auto covariance matrix of pixels included in an area having a predetermined size in an image to be interpolated,

Represents a cross covariance matrix of pixels included in a region having a predetermined size. According to an embodiment of the present invention, R and < RTI ID = 0.0 >

May be the same as the size of the region where the number of edges is counted by the edge region classifying unit 13. [

는 고해상도의 이미지에서 계산할 수 없다. 하지만, 통계적으로 보간 후의 고해상도 이미지는 보간 전의 저해상도 이미지와 기하학적으로 동일성(geometric duality)을 가지므로, 상기 가중치 α를 계산하기 위해 필요한 R 및

는 보간 전의 저해상도 이미지로부터 계산될 수 있다.Since there is no information on the pixel values Y _{2i +1 and 2j + 1} of the pixel 50 at the time of calculating the weight value a,

Can not be calculated from high resolution images. However, statistically, the high-resolution image after interpolation has geometric duality with the low-resolution image before interpolation, so that R and < RTI ID = 0.0 >

Can be calculated from the low resolution image before interpolation.

According to another embodiment of the present invention, the weight α may be calculated based on the distance between the pixel 50 to be interpolated and the adjacent pixels 501 to 512. For example, the weight a may be calculated to be inversely proportional to the distance between the pixel 50 to be interpolated and the adjacent pixels 501 to 512, or may be calculated to be inversely proportional to the square of the distance.

The interpolator 14 may use less than four adjacent pixels located around the pixel to interpolate the pixels contained in the second area 201b. In other words, the interpolator 14 interpolates the pixels using more than four adjacent pixels for the first section 201a that includes the primary edge contributing to the main contour of the image 20, The pixel can be interpolated using fewer than four adjacent pixels for the second region 201b including the sub-edges contributing to the detail or texture of the pixel. According to one embodiment of the present invention, the interpolator 14 can perform interpolation using a linear interpolation scheme for pixels of non-edge regions 202-205 that do not include edges in the image 20 have.

As a result, the image interpolating apparatus 100 according to an embodiment of the present invention is capable of interpolating pixels in a plurality of directions without referring to only pixels in a specific direction, with respect to a long edge that greatly affects the sharpness and image quality of the image 20. [ So that a more accurate and accurate image can be obtained. Furthermore, the image interpolating apparatus 100 can reduce the amount of computation required for interpolation by referring to a smaller number of pixels than a long edge, for a short edge that has less influence on the sharpness and image quality of the image 20. [

8, the interpolator 14 uses 12 adjacent pixels 501-512 to interpolate the pixels 50 included in the first section 201a, 12, the interpolator 14 includes six adjacent pixels (501 to 506 in Fig. 9), eight adjacent pixels (501 to 506, 511 and 512 in Fig. 10) located around the pixel 50, (501 to 506, 507 and 509 in Fig. 11; 501 to 504 and 507 to 510 in Fig. 12), ten adjacent pixels (501 to 510 in Fig. Further, according to an embodiment, the interpolator 14 may perform interpolation with reference to five, seven, nine and eleven adjacent pixels located around the pixel 50, or may reference more than twelve adjacent pixels Interpolation may be performed.

FIG. 14 is a flowchart illustrating an image interpolation method according to an embodiment of the present invention. The image interpolation method according to an embodiment of the present invention can perform interpolation in a plurality of ways with respect to an edge area including an edge in an image using the image interpolation apparatus 100 described above.

As shown in FIG. 14, an image interpolation method 200 according to an exemplary embodiment of the present invention includes a step S21 of identifying an edge region from an image, a step S22 of classifying the edge region into a plurality of regions, , And performing interpolation (S23) in different ways for each zone.

The step S21 may be performed by the edge region identifying unit 11 included in the image interpolating apparatus 100. [ The edge region identification section 11 can identify an edge region 201 including an edge and a non-edge region 202-205 that does not include an edge in the image 20 to be interpolated.

15 is a flow chart illustrating step S21 of identifying an edge region from an image in accordance with an embodiment of the present invention. As shown in FIG. 15, according to an embodiment of the present invention, the step S21 of identifying an edge region from the image includes a step S211 of loading an area of a predetermined size from the image 20, (S212) of calculating the variance of the pixel values of the pixels included in the on-area, and if the variance of the pixel values is larger than the predetermined threshold (YES in (S213)), (Step S214).

According to one embodiment, the size of the area to be loaded may be 4 X 4 to calculate the variance of the pixel values, but the size of the area is not limited thereto and may be smaller or larger than 4 X 4.

According to an embodiment of the present invention, the step S21 of identifying an edge region from the image may be performed when the variance of pixel values included in the retrieved region is smaller than or equal to the threshold value (NO in (S213) And determining (S215) the ON region as the non-edge region 202-205.

The step S22 may be performed by the edge region classifying unit 13 included in the image interpolating apparatus 100. [ The edge region classifying section 13 may classify the edge region 201 identified from the image 20 into a plurality of regions.

16 is a flow chart illustrating the step of classifying an edge region into a plurality of regions according to an embodiment of the present invention. As shown in FIG. 16, according to an embodiment of the present invention, the step S22 of classifying the edge region into a plurality of regions may include generating an edge map by detecting an edge included in the edge region 201 A step S222 of calculating a number of edges included in the loaded area, a step S222 of calculating a number of edges included in the loaded area, a step of calculating a number of edges included in the loaded area, (S224). If the number of edges is larger than the threshold value (YES in S224) (S224: No), the step of determining the loaded zone as the first zone (S225) (Step S226).

The step (S221) of generating the edge map may be performed by the edge map generating unit 12 included in the image interpolating apparatus 100. [ According to an embodiment of the present invention, the step of generating the edge map (S221) includes the steps of: detecting a first pixel corresponding to an edge in the edge region 201; detecting a plurality of pixels located around the first pixel Detecting a second pixel corresponding to an edge of the first pixel, and connecting the first pixel and the second pixel.

The process of detecting a pixel corresponding to the edge can be performed until a new edge pixel is no longer detected around the edge pixel. When an edge is detected in the edge area 201 in this way, the edge can be indexed with a predetermined number, symbol or character. Then, the edge map generation unit 12 detects pixels corresponding to other edges in the edge region 201, detects pixels corresponding to the edges of the surrounding pixels with the detected pixel as a starting point, , Another edge can be detected.

According to one embodiment of the present invention, the size of the region to be fetched from the edge map for calculating the number of edges may be 15 X 15, but the size of the region is not limited thereto and may be set to be smaller or larger than 15 X 15 have.

If the number of edges in the loaded area is less than or equal to a predetermined threshold value, the edge area classifying unit 13 may determine the corresponding area as the first area, that is, the main area 201a. If the number of edges in the loaded area is greater than the threshold value, the edge area classifying unit 13 may determine the area as the second area, that is, the sub-area 201b.

If the number of edges included in the loaded area is below the threshold, then the edges included in the area may be considered relatively long, and the area containing these long edges may be classified as the main area contributing to the main contour of the image . Conversely, if the number of edges included in the loaded area exceeds the threshold, the edges included in the area may be considered to be relatively short in length, and the area containing such short edges may not contribute to the main contour of the image, Or sub-regions that contribute to the texture representation of the object.

According to one embodiment, the threshold may be set to any one of 1 to 5, but is not limited thereto and may be set to a value larger than 5 according to an image to be interpolated.

The step S23 of performing the interpolation in different ways for each of the zones may be performed by the interpolating unit 14 included in the image interpolating apparatus 100. [ According to an embodiment of the present invention, performing the interpolation (S23) in different ways for each of the zones may include interpolating pixels included in the first zone 201a by more than four And using the number of adjacent pixels. In addition, the step of performing the interpolation S23 may include using less than four adjacent pixels located around the pixel to interpolate the pixels included in the second area 201b.

In other words, the number of contiguous pixels used to interpolate the main zone 201a contributing to the main contour of the image 20 may be used to interpolate the sub-zone 201b that contributes to the detail or texture of the image 20 May be greater than the number of adjacent pixels used.

According to one embodiment, performing the interpolation (S23) may include using 12 adjacent pixels (501 to 512) to interpolate pixels contained in the first region (201a) .

According to one embodiment, the step of performing the interpolation S23 may comprise interpolating pixels of the non-edge areas 202 to 205 that do not include edges in the image 20 in a linear interpolation manner have.

17 is a block diagram illustrating an image processing apparatus according to an embodiment of the present invention. 17, the image processing apparatus 300 according to an exemplary embodiment of the present invention may include a receiving unit 31, a processing unit 32, a storage unit 33, and an output unit 34.

The receiving unit 31 may receive an image. The image is an original image to be interpolated. The image processing apparatus 300 according to an embodiment of the present invention interpolates new pixels using pixels included in the original image, thereby enlarging the size of the original image, Images can be converted to high-resolution images.

The processing unit 32 may process the received image. According to an embodiment of the present invention, the processing unit 32 may include an edge region identifying unit 321 for identifying an edge region including an edge from the received image, an edge region detecting unit 321 for detecting an edge included in the edge region, An edge region classifying unit 323 for classifying the edge region into a plurality of regions by using the edge map, and an interpolating unit 324 for performing interpolation in different ways for each region .

The edge region identifying unit 321, the edge map generating unit 322, the edge region classifying unit 323 and the interpolating unit 324 included in the processing unit 32 are connected to the image interpolator 100 The edge map generating section 12, the edge region classifying section 13, and the interpolating section 14 included in the edge region identifying section 11, The processing unit 32 may be a processor including a CPU, a GPU, and the like for performing operations for image processing.

The storage unit 33 may store data used for image or image processing. The storage unit 33 may be a storage device that stores predetermined data, and may include, for example, a RAM, a ROM, a hard disk, an SSD, or the like.

The output unit 34 may output the processed image. The output unit 34 may transmit the processed image to a display device such as an LCD or a PDP so that the image is displayed on the screen of the display device.

According to an embodiment of the present invention, if the number of edges included in one region of the edge region 201 is less than or equal to a predetermined threshold, the edge region classifying unit 323 divides the region into the first region 201a, And if the number of edges is larger than the threshold value, the zone can be determined as the second zone 201b. According to an embodiment, the threshold value may be set to two or more, and the edge region classifying unit 323 may classify the edge region 201 into three or more regions.

The edge region classifying unit 323 classifies edge regions according to the number of edges included in one region. However, according to another embodiment, the edge region classifying unit 323 classifies the edge regions included in one region The edge region may be classified according to the length of the longest edge. For example, when the length of the longest edge among the edges included in one region is longer than a preset threshold value, the edge region classifying unit 323 classifies the corresponding region into the first region 201a, Is shorter than or equal to the threshold value, the zone may be classified into the second zone 201b.

According to another embodiment, the edge region classifying section 323 may classify the edge region according to the entire length of edges included in one region. For example, the edge region classifying unit 323 sums the lengths of the edges included in one region and classifies the regions into the first region 201a if the length of the summed edge is longer than a preset threshold value, If the length of the edge is shorter than or equal to the threshold, the zone may be classified as the second zone 201b.

According to one embodiment of the present invention, the interpolator 324 can perform interpolation for pixels included in the first section 201a using more than four adjacent pixels located around the pixel . In addition, the interpolator 324 may interpolate pixels included in the second region 201b using four or less adjacent pixels located around the pixel. The interpolator 324 may also perform interpolation using a linear interpolation scheme for pixels of the non-edge regions 202-205 that do not include edges in the image 20. [ In other words, the interpolator 324 can use more adjacent pixels than the adjacent pixels used for pixel interpolation in the second region 201b, for pixel interpolation in the first region 201a.

According to one embodiment, the interpolator 324 may use twelve adjacent pixels for interpolation of the first section 201a, but it is not limited thereto and may use five to eleven adjacent pixels, or more than twelve Many adjacent pixels may be used.

According to one embodiment, the interpolator 324 may use a non-linear interpolation scheme to interpolate the second section 201b.

The image interpolation method 200 according to an embodiment of the present invention may be stored in a computer-readable recording medium that is manufactured as a program to be executed in a computer. The computer-readable recording medium includes all kinds of storage devices in which data that can be read by a computer system is stored. Examples of the computer-readable recording medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like.

In the above, an image interpolating apparatus, an image interpolating method, and an image processing apparatus using the same are described in which an edge region including an edge at the time of interpolation of an image is classified into a plurality of regions, and interpolation is performed by applying different interpolation methods to each region Respectively. The image interpolating apparatus, the image interpolating method, and the image processing apparatus may increase the number of adjacent pixels used for pixel interpolation for a main region including a relatively long edge and contribute to the main contour of the image, For subregions that do not significantly affect the main contour of the image, including short edges, the number of adjacent pixels used for pixel interpolation can be reduced.

As a result, the image interpolating apparatus, the image interpolating method, and the image processing apparatus can relatively accurately and vigorously recover the long edges of the image, thereby obtaining an image in which the image quality is not significantly degraded as compared with the original image even after interpolation. Also, the image interpolating apparatus, the image interpolating method, and the image processing apparatus set the number of adjacent pixels used for interpolation of pixels to be different for each region of the edge region, thereby preventing the amount of computation required for interpolation from rapidly increasing .

100: image interpolating device 11: edge area identifying unit
12: edge map generating unit 13: edge region classifying unit
14: interpreter

Claims (25)

An edge region identification unit for identifying an edge region including an edge from the image;
An edge map generation unit for generating an edge map by detecting an edge included in the edge region;
An edge region classifying unit classifying the edge region into a plurality of regions; And
An interpolation section which performs interpolation in a different manner for each of the zones;
/ RTI >
Wherein the edge region classifying unit classifies the edge region according to the number of edges included in one region of the edge map. The method according to claim 1,
Wherein the edge region identification unit identifies the edge region by calculating a variance of pixel values of pixels included in one region of the image. 3. The method of claim 2,
Wherein the edge region identification unit comprises:
If the variance of the pixel value is greater than a preset threshold value,
And determines the one region as a non-edge region if the variance of the pixel values is less than or equal to the threshold value. delete The method according to claim 1,
Wherein the edge map generator comprises:
Detecting a first pixel corresponding to an edge in the edge region, detecting a second pixel corresponding to an edge of a plurality of pixels located around the first pixel, and connecting the first pixel and the second pixel Image interpolation device. delete The method according to claim 1,
Wherein the edge region classifier comprises:
If the number of edges included in the one zone is less than or equal to a preset threshold value, the one zone is determined as the first zone,
Wherein if the number of edges included in the one area is greater than the threshold value, the one area is determined as the second area. 8. The method of claim 7,
Wherein the interpolator uses more than four adjacent pixels located around the pixel to interpolate pixels contained in the first region. 9. The method of claim 8,
Wherein the interpolator interpolates the pixel using twelve adjacent pixels surrounding the pixels included in the first zone. 9. The method of claim 8,
Wherein the interpolator calculates a pixel value of the pixel by applying a weight to pixel values of each of the plurality of adjacent pixels, 11. The method of claim 10,
Wherein the weight is calculated using a Wiener filter. 11. The method of claim 10,
Wherein the weight is calculated based on a distance between the pixel and the adjacent pixel. 8. The method of claim 7,
Wherein the interpolator uses less than four adjacent pixels located around the pixel to interpolate pixels contained in the second region. The method according to claim 1,
Wherein the interpolator uses a linear interpolation scheme to interpolate pixels in a non-edge region that do not include an edge in the image. delete Identifying an edge region comprising an edge from the image;
Detecting an edge included in the edge region to generate an edge map;
Classifying the edge region into a plurality of regions according to the number of edges included in one region of the edge map; And
Performing interpolation in different ways for each of the zones;
/ RTI > 17. The method of claim 16,
Wherein identifying the edge region comprises:
Loading an area of predetermined size from the image;
Calculating a variance of pixel values of pixels included in the loaded area; And
Determining the read area as an edge area if the variance of the pixel value is greater than a preset threshold;
/ RTI > 17. The method of claim 16,
Wherein classifying the edge region into a plurality of zones comprises:
Loading a zone of predetermined size from the edge map;
Calculating the number of edges included in the loaded area;
Determining the loaded zone as a first zone if the number of edges is less than or equal to a predetermined threshold; And
Determining the loaded zone as a second zone if the number of edges is greater than a predetermined threshold;
/ RTI > 19. The method of claim 18,
Wherein performing the interpolation comprises:
Interpolating the pixel using more than four adjacent pixels located around a pixel included in the first region. 19. The method of claim 18,
Wherein performing the interpolation comprises:
Interpolating the pixel using less than four adjacent pixels located around a pixel included in the second region. 17. The method of claim 16,
Wherein performing the interpolation comprises:
Interpolating pixels of a non-edge region that does not include an edge in the image in a linear interpolation manner. A receiving unit for receiving an image;
A processing unit for processing the received image;
A storage unit for storing data used for image or image processing; And
And an output unit for outputting the processed image,
Wherein the processing unit comprises:
An edge region identification unit for identifying an edge region including an edge from the received image;
An edge map generation unit for generating an edge map by detecting an edge included in the edge region;
An edge region classifying unit classifying the edge region into a plurality of regions according to the number of edges included in one region of the edge map using the edge map; And
An interpolation section which performs interpolation in a different manner for each of the zones;
And the image processing apparatus. 23. The method of claim 22,
Wherein the edge region classifier comprises:
If the number of edges included in one region of the edge region is less than or equal to a predetermined threshold value, the region is determined as the first region,
And determines the zone as a second zone if the number of edges included in one zone of the edge zone is larger than the threshold value. 24. The method of claim 23,
Wherein the interpolator comprises:
Performing interpolation for pixels included in the first zone using more than four adjacent pixels located around the pixel,
Performing interpolation with respect to the pixels included in the second zone using four or less adjacent pixels located around the pixel,
And performs interpolation in a linear interpolation manner on pixels of a non-edge region that does not include edges in the received image. A computer-readable recording medium,
A recording medium on which a program for implementing the image interpolation method according to any one of claims 16 to 21 is recorded.

Images