JP2004056586A - Image processor and image processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To generate a smooth interpolating image with few jaggies for an original image having an oblique direction component by adaptively changing a selecting method of a reference pixel. <P>SOLUTION: An edge detecting means 2 detects pixels constituting an edge part of an image in the original image by an interlaced signal 1. A reference pixel selecting means 3 selects a reference pixel for generating an interpolation pixel according to the arrangement pattern of the detected pixels constituting the edge part of the image within a prescribed area centered on an attentional interpolation pixel. An interpolation pixel generating means 4 generates an interpolation pixel on the basis of the selected reference pixel and outputs a progressive signal 5 constituting an interpolation image. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an image processing apparatus and an image processing method for generating a progressive signal by performing pixel interpolation processing on an interlaced signal.
[0002]
[Prior art]
In order to display an interlaced signal used in television on a computer display, it is necessary to perform pixel interpolation processing of the interlaced signal every other line and convert it into a progressive signal used in a computer. FIG. 7 is a diagram for explaining pixel interpolation processing in the same field of an interlace signal in a conventional image processing apparatus. In the figure, lines L and L + 1 are original lines of an interlace signal, and an interpolation pixel is generated from pixels on the original line.
[0003]
As shown in FIG. 7, let the pixel values of each line be A (X-1), A (X), A (X + 1) and B (X-1), B (X), B (X + 1). Here, when calculating the pixel value P (X) of the interpolation pixel, pixels that are vertically adjacent to each other on the upper and lower lines of the interpolation pixel to be determined are used as reference pixels, that is, the pixel having the pixel value A (X) and the pixel value B (X ) Is used as a reference pixel, an average value of the pixel values of this reference pixel pair is determined, and the average value is used as the pixel value of the interpolation pixel. That is,
P (X) = (A (X) + B (X)) / 2
And Hereinafter, this interpolation method is referred to as a vertical interpolation method.
[0004]
This vertical interpolation method has an advantage that the circuit configuration and the like are small.However, when an original image based on an interlaced signal includes a diagonal line, the jaggies in which an oblique component of an interpolated image based on a progressive signal becomes jagged are increased. There is.
[0005]
As another interpolation method, there is a correlation interpolation method in which a pair having a small difference in pixel value is selected as a reference pixel for generating a target interpolation pixel. In this method, the absolute value | A (X−) of the difference between the pixel having the pixel value A (X−1) and the pixel having the pixel value B (X + 1) that is point-symmetric with respect to the interpolation pixel having the pixel value P (X). 1) Find −B (X + 1) |. Hereinafter, similarly, | A (X) -B (X) | and | A (X + 1) -B (X-1) | are obtained.
[0006]
The smallest one of the absolute values of the three differences is regarded as having the highest correlation, and this pixel is set as a reference pixel necessary for generating an interpolation pixel. For example, when | A (X−1) −B (X + 1) | is the minimum, the pixel having the pixel value A (X−1) and the pixel having the pixel value B (X + 1) are set as a reference pixel pair, and the pixel of the interpolation pixel is set. Value P (X)
P (X) = (A (X-1) + B (X + 1)) / 2
And However, in the case of this correlation interpolation method, if an appropriate reference pixel pair cannot be selected by mistake, image degradation may be caused.
[0007]
[Problems to be solved by the invention]
Since the conventional image processing apparatus is configured as described above, in the case of the vertical interpolation method, when the original image includes diagonal lines, the jaggies in which the diagonal components of the interpolated image are jagged increase, and However, in the case of the correlation interpolation method, if an appropriate reference pixel pair cannot be selected, there is a problem that image deterioration is caused.
[0008]
The present invention has been made to solve the above-described problem, and adaptively changes a method of selecting a reference pixel between a case where an interpolation pixel is generated near an edge portion of an image and a case where the interpolation pixel is not provided. It is another object of the present invention to provide an image processing apparatus and an image processing method capable of generating a smooth interpolated image with less jaggies for an original image having an oblique component.
[0009]
[Means for Solving the Problems]
An image processing apparatus according to the present invention includes an edge detection unit configured to detect a pixel constituting an edge portion of an image in an original image based on an interlaced signal, and an image detection device configured to detect a detected image within a predetermined region centered on an interpolation pixel of interest. Reference pixel selecting means for selecting a reference pixel for generating an interpolation pixel according to an arrangement pattern of pixels forming an edge portion, and a progressive signal for generating an interpolation pixel based on the selected reference pixel and forming an interpolation image And an interpolation pixel generation means for outputting
[0010]
In the image processing apparatus according to the present invention, the edge detecting means compares a difference between horizontally adjacent pixels with a predetermined threshold value, and determines a pixel having a difference larger than the threshold value as a pixel constituting an edge portion of the image. Is detected.
[0011]
In the image processing apparatus according to the present invention, the reference pixel selection unit refers to the pixels constituting the edge portion of the image detected in the area composed of the left and right ± 2 pixels on the upper and lower ± 1 line around the interpolation pixel. The pixel is selected as the pixel, and the interpolated pixel generation means generates an interpolated pixel having the average value of the pixel values of the reference pixels as the pixel value.
[0012]
In the image processing apparatus according to the present invention, the reference pixel selecting means selects a pixel which is point-symmetric with respect to the interpolation pixel as a reference pixel.
[0013]
In the image processing apparatus according to the present invention, the reference pixel selection means selects two pairs of pixels which are point-symmetric with respect to the interpolation pixel as reference pixels.
[0014]
In the image processing apparatus according to the present invention, the reference pixel selection unit selects a pixel that is not point-symmetric with respect to the interpolation pixel as the reference pixel, and the interpolation pixel generation unit weights the pixel value of the reference pixel to obtain an average value Things.
[0015]
In the image processing apparatus according to the present invention, when the edge detecting means does not detect a pixel forming an edge portion of the image, the reference pixel selecting means selects a pixel immediately above the one line above the interpolation pixel and a line below the interpolated pixel. The pixel immediately below is selected as a reference pixel, and the interpolation pixel generation means generates an interpolation pixel having an average value of the pixel values of the reference pixels as a pixel value.
[0016]
An image processing method according to the present invention includes a first step of detecting pixels constituting an edge portion of an image in an original image based on an interlaced signal, and a method of detecting an image detected within a predetermined region centered on an interpolation pixel of interest. A second step of selecting a reference pixel for generating an interpolated pixel according to an arrangement pattern of pixels constituting an edge portion of the image, and a progressive step of generating an interpolated pixel based on the selected reference pixel and forming an interpolated image And a third step of outputting a signal.
[0017]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, an embodiment of the present invention will be described.
Embodiment 1 FIG.
FIG. 1 is a block diagram showing a configuration of an image processing apparatus according to Embodiment 1 of the present invention. In the figure, 1 is an interlace signal constituting an input original image, 2 is an edge detecting means for detecting a pixel constituting an edge portion of the image in the original image, and 3 is a predetermined area around an interpolation pixel of interest. The reference pixel selecting means 4 for selecting a reference pixel for generating an interpolated pixel is selected by the reference pixel selecting means 3 in accordance with the arrangement pattern of the pixels constituting the edge portion of the image detected by the edge detecting means 2. An interpolated pixel generation unit that generates an interpolated pixel based on the reference pixel and outputs a progressive signal 5 that constitutes an interpolated image.
[0018]
Next, the operation will be described.
The edge detecting means 2 detects pixels constituting an edge portion of the image in the original image based on the input interlace signal 1. Here, the edge part of the image is detected by calculating the first derivative of the adjacent pixel.
[0019]
2A and 2B are diagrams for explaining detection of an edge portion of an image by the edge detecting means 2. FIG. 2A shows a pixel value a (t) with respect to time t, and FIG. The difference absolute value DIFF (t) from the pixel value is shown. Assuming that the pixels input at times t and t + 1 are a (t) and a (t + 1), respectively, the edge detection means 2 calculates the absolute value DIFF (t) of the difference between the pixel values by the following equation.
DIFF (t) = | a (t + 1) -a (t) |
[0020]
The edge detecting means 2 compares the obtained absolute value DIFF (t) with a predetermined threshold TH, and if DIFF (t) is larger than the threshold TH, the edge detector 2 determines a pixel having a pixel value a (t) of the image. If the pixel is detected as a pixel constituting an edge portion and DIFF (t) is smaller than the threshold value TH, a pixel having a pixel value a (t) is determined as a pixel not constituting an edge portion of the image. In the example of FIG. 2, the edge detecting means 2 detects a pixel having a pixel value a (3) as a pixel constituting an edge of an image.
[0021]
The reference pixel selecting means 3 selects a reference pixel pair necessary for obtaining a target interpolation pixel. FIG. 3 is a diagram showing the interpolation pixel of interest and the pixels on the original line. In the figure, the mark ◎ indicates the interpolation pixel P of interest, the upper original line of the interpolation pixel P is line A, the lower original line is line B, and the pixels on line A are a, b, c, d. , E, and the pixels on the line B are p, q, r, s, and t.
[0022]
FIG. 4 is a diagram for explaining selection of a reference pixel pair by the reference pixel selection means 3 when an edge portion of an image exists in a predetermined area. In the figure, the pixels constituting the edge portion of the image detected by the edge detecting means 2 are indicated by ●, and the other pixels are indicated by ○. The reference pixel selection means 3 selects a reference pixel for generating an interpolation pixel by an arrangement pattern of pixels constituting an edge portion of the image detected by the edge detection means 2 within a predetermined area centered on the interpolation pixel of interest. In order to make the circuit scale of the image processing device appropriate, an area including left and right ± 2 pixels on the upper and lower ± 1 line centering on the interpolation pixel is set as the predetermined area.
[0023]
In the example of FIG. 4, the pixel b on the line A and the pixel s on the line B are detected as the pixels constituting the edge portion of the image. The edge portion of the image indicated by the dotted line exists in the area composed of the left and right ± 2 pixels, and the reference pixel selection unit 3 selects the pixel b and the pixel s as the reference pixels of the interpolation pixel P.
[0024]
The interpolation pixel generation means 4 generates an interpolation pixel P from the selected reference pixel b and reference pixel s. That is, the interpolated pixel generation means 4 generates an interpolated pixel P having the average value of the pixel value of the reference pixel b and the pixel value of the reference pixel s as the pixel value, and outputs the progressive signal 5 constituting the interpolated image.
[0025]
FIG. 5 is a diagram illustrating selection of another reference pixel pair by the reference pixel selection unit 3 when an edge portion of an image exists in a predetermined area. The example of FIG. 4 illustrates an example in which the pixel b and the pixel s located at point-symmetric positions with respect to the interpolation pixel P of interest are detected as the pixels forming the edge portion of the image. Shows an example in which the pixel b and the pixel t which are not point-symmetric with respect to the interpolation pixel P of interest are detected as pixels constituting an edge portion of the image.
[0026]
When a pixel at a position as shown in FIG. 5 is detected as a pixel constituting an edge portion of the image, the reference pixel selection means 3 selects the pixel b as a reference pixel of the interpolation pixel P and replaces the pixel t with the pixel t. The pixel s located at a point symmetrical position with the pixel b is simply selected. Then, as in the case shown in FIG. 4, the interpolated pixel generation means 4 generates an interpolated pixel P having an average pixel value of the pixel value of the reference pixel b and the average value of the pixel values of the reference pixel s to form an interpolated image. A progressive signal 5 is output. Here, the pixel a located at a point symmetrical position with respect to the pixel t may be selected.
[0027]
When a pixel at a position as shown in FIG. 5 is detected as a pixel constituting an edge portion of the image, the reference pixel selection unit 3 selects the pixel b and the pixel t as reference pixels of the interpolation pixel P, When calculating the average value of the pixel value of the reference pixel b and the pixel value of the reference pixel t as the pixel value of the interpolation pixel P, the interpolation pixel generation unit 4 determines the pixel value of the reference pixel t at a position away from the interpolation pixel P. May be made lighter than the reference pixel b to obtain the pixel value of the interpolated pixel P, and an interpolated image with little image degradation can be generated even for an original image including a diagonal line that changes rapidly.
[0028]
Further, when a pixel at a position as shown in FIG. 5 is detected as a pixel constituting an edge portion of the image, the reference pixel selecting means 3 sets a point symmetrical position with respect to the interpolation pixel P as a reference pixel of the interpolation pixel P. , The pixel a and the pixel t, and the pixel b and the pixel s are selected, and the interpolation pixel generation unit 4 calculates the average value of the pixel values of the reference pixels a, b, s, and t by using the interpolation pixel P Similarly, an interpolated image with little image degradation can be generated even for an original image including a diagonal line that changes rapidly.
[0029]
FIG. 6 is a view for explaining selection of a reference pixel pair by the reference pixel selection means 3 when an edge portion of an image does not exist in a predetermined area. As described above, when there is no pixel constituting the edge portion of the image in the area composed of the left and right ± 2 pixels on the upper and lower ± 1 line around the interpolation pixel P of interest, the reference pixel selection means 3 The pixel c and the pixel r are selected as reference pixels of the interpolation pixel P by the vertical interpolation method.
[0030]
The interpolation pixel generation means 4 generates an interpolation pixel P from the selected reference pixel c and reference pixel r. That is, the interpolated pixel generation means 4 generates an interpolated pixel P having an average value of the pixel value of the reference pixel c and the pixel value of the reference pixel r as a pixel value, and outputs the progressive signal 5 constituting the interpolated image.
[0031]
As described above, the reference pixel selection unit 3 determines, from the pixels in the area of ± 2 pixels on the left and right ± 1 lines around the interpolation pixel P of interest, according to the arrangement pattern of the pixels constituting the edge portion of the image. A reference pixel pair is selected, and the interpolated pixel generation unit 4 generates an interpolated pixel P having a pixel value of an average value or an average value adjusted by weighting, based on the reference pixel pair obtained by the reference pixel selection unit 3.
[0032]
As described above, according to the first embodiment, oblique interpolation is performed by the correlation interpolation method in a region where an edge portion of an image exists, and a vertical interpolation method is used in a region where an edge portion of the image does not exist. By doing so, it is possible to suppress the deterioration of the image due to the selection of the wrong reference pixel pair, and it is possible to generate a smooth interpolated image with less jaggies with respect to the original image having an oblique component. Can be
[0033]
Further, according to the first embodiment, the reference pixel selecting means 3 determines the edge portion of the image from the pixels in the area of ± 2 pixels on the upper and lower ± 1 line around the interpolation pixel P of interest. By selecting the reference pixel pair according to the arrangement pattern of the constituent pixels, an effect is obtained that the circuit scale of the image processing apparatus can be made appropriate.
[0034]
Further, according to the first embodiment, the reference pixel selection unit 3 selects a pixel that is not point-symmetric with respect to the interpolation pixel as a reference pixel, and the interpolation pixel generation unit 4 weights the pixel value of the reference pixel. By obtaining the average value, it is possible to obtain an effect that an interpolated image with little image deterioration can be generated even for an original image including a diagonal line that changes rapidly.
[0035]
【The invention's effect】
As described above, according to the present invention, it is possible to suppress deterioration of an image due to selection of an incorrect reference pixel pair, and to perform smooth interpolation with less jaggies with respect to an original image having an oblique component. Can be generated.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration of an image processing device according to a first embodiment of the present invention.
FIG. 2 is a diagram for explaining detection of an edge portion of an image by an edge detecting means in the image processing device according to the first embodiment of the present invention;
FIG. 3 is a diagram showing an interpolation pixel of interest and a pixel on an original line in the image processing apparatus according to the first embodiment of the present invention;
FIG. 4 is a diagram illustrating selection of a reference pixel pair by a reference pixel selection unit when an edge portion of an image exists in a predetermined area in the image processing device according to the first embodiment of the present invention;
FIG. 5 is a diagram illustrating selection of another reference pixel pair by a reference pixel selection unit when an edge portion of an image exists in a predetermined area in the image processing device according to the first embodiment of the present invention;
FIG. 6 is a diagram illustrating selection of a reference pixel pair by a reference pixel selection unit when an edge portion of an image does not exist in a predetermined area in the image processing device according to the first embodiment of the present invention;
FIG. 7 is a diagram illustrating pixel interpolation processing in the same field of an interlaced signal in a conventional image processing apparatus.
[Explanation of symbols]
1 interlace signal, 2 edge detection means, 3 reference pixel selection means, 4 interpolation pixel generation means, 5 progressive signal.

Claims (8)

In an image processing device that interpolates an interlaced signal and converts it to a progressive signal,
Edge detection means for detecting pixels constituting an edge portion of the image in the original image by the interlace signal,
A reference pixel for selecting a reference pixel for generating the interpolation pixel in a predetermined area centered on the interpolation pixel of interest, based on an arrangement pattern of pixels constituting an edge portion of the image detected by the edge detection means. Selecting means;
An image processing apparatus comprising: an interpolation pixel generation unit that generates the interpolation pixel based on the reference pixel selected by the reference pixel selection unit and outputs the progressive signal that forms the interpolation image. The edge detecting means compares a difference between horizontally adjacent pixels and a predetermined threshold value, and detects a pixel whose difference is larger than the threshold value as a pixel constituting an edge portion of the image. The image processing apparatus according to claim 1. The reference pixel selecting means selects, as a reference pixel, a pixel constituting an edge portion of an image detected in a region consisting of ± 2 pixels on the left and right on the ± 1 line above and below the interpolation pixel,
3. The image processing apparatus according to claim 2, wherein the interpolated pixel generation means generates an interpolated pixel having an average value of the pixel values of the reference pixels as a pixel value. 4. The image processing apparatus according to claim 3, wherein the reference pixel selecting means selects a pixel which is point-symmetric with respect to the interpolation pixel as a reference pixel. 5. The image processing apparatus according to claim 4, wherein the reference pixel selection unit selects two pairs of pixels that are point-symmetric with respect to the interpolation pixel as reference pixels. The reference pixel selecting means selects a pixel that is not point-symmetric with respect to the interpolation pixel as a reference pixel,
4. The image processing apparatus according to claim 3, wherein the interpolated pixel generation means obtains an average value by weighting the pixel value of the reference pixel. When the edge detecting means does not detect the pixels constituting the edge portion of the image,
The reference pixel selection means selects a pixel immediately above one line of the interpolation pixel of interest and a pixel immediately below one line below as a reference pixel,
2. The image processing apparatus according to claim 1, wherein the interpolated pixel generation means generates the interpolated pixel having an average value of the pixel values of the reference pixels as a pixel value. In an image processing method of interpolating an interlace signal and converting it to a progressive signal,
A first step of detecting pixels constituting an edge portion of the image in the original image based on the interlace signal,
In a predetermined area centered on the interpolated pixel of interest, a reference pixel for generating the interpolated pixel is selected by an arrangement pattern of pixels constituting an edge portion of the image detected in the first step. Two steps,
Generating the interpolated pixel based on the reference pixel selected in the second step, and outputting the progressive signal forming the interpolated image.

Images