JPH0823509A - Field interpolation method for video signal - Google Patents

Abstract

PURPOSE:To provide the field interpolation method of video signals capable of improving the interpolation ability of pictures provided with a shallow inclination line and a thin inclination line without increasing the number of reference picture element data so much. CONSTITUTION:In an interpolation filtering processing for extracting two picture elements present at the same horizontal position on upper and lower adjacent two lines within the same field of the line provided with a point to be generated by interpolation and at least one each of the picture elements at the positions respectively successive to the right and left of the two picture elements as reference picture elements, plural data sets composed of the respective data of more than two points of the reference picture elements included with more than one data sets composed of the respective data of more than three points of the reference picture elements are prepared, one data set is selected as an interpolation data set from the information of the plural data sets and the interpolation data of all the points to be interpolated are decided from the selected interpolation data set.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a video signal field interpolation method, and more particularly, in a video terminal signal processing technique, a video signal for generating a frame image from a single field image stored in a memory by using an interpolating means. The present invention relates to the field interpolation method of.

[0002]

2. Description of the Related Art Generally, in a video printer for printing an image from a video terminal signal, only an image for one field is stored in an image memory and a frame image is obtained by interpolating the other field. There is. Density data A to F of field pixels as reference pixels are arranged as shown in FIG. 12, and density data X to Z of interpolation pixels are interpolated therebetween. There are various interpolating methods as follows.

The densities X to Z of the interpolated pixels are respectively set to A, B, and
The interpolation method of C or D, E, and F and the densities X to Z of the interpolated pixels are (A + D) / 2, (B + E) / 2,
There is an interpolation method of (C + F) / 2. In addition, JP-A-4
-347985 and "Japan Hardco
py'91 "," A-F |, | ", as disclosed in" Image Processing Techniques for Consumer Video Printers "
B-E |, | C-D |, an interpolation method for extracting a set having a minimum difference and selecting the average value of the extracted set as the density Y of the interpolated pixel, and the Institute of Electronics, Information and Communication Engineers Technique 19
91, 91-70271, "High-quality image processing for video printers", the six values of the density data A to F of the field pixels are binarized data, and the result and the pattern of FIG. The direction to be interpolated is determined by pattern matching using a table, and (A + F) / 2, (B + E) / 2, (C
There is an interpolation method that selects either + D) / 2. Further, as disclosed in "Japan Hardcopy '90", "High-quality image processing in color video printers", A and B, B and C, D and E, E and F, A and D, and C The F direction is compared, and the result is compared with the pattern matching using the pattern table of FIG. 13 to determine the interpolation direction, and the pixel density Y of the field is (A + F) / 2, (B + E) / 2, (C + D). There is an interpolation method for selecting either / 2.

[0004]

Since the conventional interpolation method is configured as described above, when interpolating the image density data at the boundary of the shallow slope as shown in FIG.
Interpolation data as shown in FIG. 15A or 15B is obtained. Although the data at the four points on the left side of the middle row in FIG. 14 are not changed to 100 and the data at the four points on the right side are not changed to 200, the corresponding interpolation data in the middle row in FIG. The change starts from the data (107) of (1) and changes to the first data (194) on the right side,
The image density change tends to start too early and the change ends too late, causing an error in the image density. As a result, in the interpolated image, as shown by the broken line in FIG. 16, a so-called jaggy on the inclined line occurs due to the density difference. Note that the mesh cells in FIG. 16 are the range of density change due to the inclined boundaries in the frame data, the hatched cells are the range of density changes due to the inclined boundaries in the interpolation data obtained by the conventional method, and the solid line is the boundary of the image. Corresponding lines and broken lines represent lines corresponding to boundaries in the output image of the conventional method. In the image of the slanted line on the interpolation line, the slant is too shallow, and this occurs on every other interpolated line as shown in FIG. 16, which causes jaggies. Therefore, in order to solve this problem, there is an interpolation method in which the number of reference pixels is increased to prevent jaggies from occurring.
For example, referring to the field density data of a 9 × 2 matrix in the example of FIG. 14, the average value of the density data values (113) in the 1st row and 9th column and the density data values (113) in the 3rd row and 1st column is 2 The data will be interpolated in row 5 columns, and jaggies will not occur. However, in the interpolation method in which the number of reference pixels is increased, the calculation processing cannot be performed in time because there are many field density data to be referred, and therefore an arithmetic calculation processor is added or a frame memory is added to start the arithmetic processing before printing. You need to store the result. Further, in the interpolation method by the pattern matching method, the table storing the matching patterns becomes large, so that a memory for the pattern table having a large capacity is required, the circuit scale becomes large, and the apparatus cost becomes high.

Further, according to the conventional method, when interpolating the image density data of a thin inclined line, an image with a broken line is often formed.

The present invention has been made to solve the above problems, and can improve the interpolation ability of an image including a shallow slope boundary and a thin slope line without increasing the number of reference pixel data. It is an object to provide a field interpolation method for video signals.

[0007]

SUMMARY OF THE INVENTION According to the present invention, the above object is to provide an interpolation method for forming an image signal for one frame from a digital image signal for one field stored in an image memory. In the same field of a line including a point to be generated, two pixels existing at the same horizontal position on two vertically adjacent lines and at least a pair of pixels at positions contiguous to the left and right of each pixel are reference pixels. And a plurality of data sets each including data of two or more reference pixels including at least one data set of each reference pixel data of three or more points are prepared. One or more data sets are selected as an interpolation data set from the information of a plurality of data sets, and the interpolation data of all the points to be interpolated are the above-mentioned interpolation data sets. It is achieved by a field interpolation process of the video signal according to claim 1 determined.

It should be noted that the data set having the smallest data variance may be selected as the interpolation data set from the plurality of data sets.

Further, a data set having the smallest difference between the maximum value and the minimum value may be selected as the interpolation data set from the plurality of data sets.

Further, the average value of the interpolation data set may be used as the interpolation data of the point to be interpolated.

According to the present invention, the above object is to provide a point to be generated by interpolation in an interpolation method for forming an image signal for one frame from a digital image signal for one field stored in an image memory. Interpolation filter processing for extracting, as reference pixels, two pixels existing at the same horizontal position on two vertically adjacent lines and at least a pair of pixels at positions contiguous to the left and right of each pixel in the same field of the containing line When the data of the reference pixels in one of the two lines are similar to each other and the data of the reference pixels in the other line monotonically increases or monotonically decreases, interpolation is performed. The interpolated data of all the points are set as the average value of the data of two or more reference pixels in one of the two lines. It is achieved by a field interpolation process of the video signal Motomeko 5.

The data of the reference pixels in one of the two lines are similar to each other,
And the data of the reference pixel in the other line is monotonically increasing, and the above-described processing is performed in which the interpolation data is the average value of the data of the two or more reference pixels in either one of the two lines. After that, when the data of the reference pixel in the other line is monotonically decreasing, it is preferable to determine that the data is the image of the image including the thin inclined line and perform the appropriate interpolation processing.

[0013]

In the field interpolation method for a video signal according to claim 1, two pixels existing at the same horizontal position on two vertically adjacent lines in the same field of a line including a point to be generated by interpolation. And at least a pair of pixels at positions that are respectively connected to the left and right of each pixel are extracted as reference pixels, and at least two reference pixels each including at least one data set consisting of data of three or more reference pixels Prepare multiple data sets consisting of data. One or more data sets are selected as the interpolation data set from the information of the plurality of data sets, and the interpolation data of all the points to be interpolated are determined from the interpolation data set. As a result, the same effect as smoothing can be obtained, and the interpolation data having a smooth slope with the jaggy angle reduced can be obtained, and the misinterpretation of the interpolation direction can be reduced, and the interpolation data with less error can be obtained.

In the field interpolation method of the video signal according to the present invention, an appropriate interpolation data set is selected from the plurality of data sets by selecting the data set having the smallest data variance as the interpolation data set. can do.

In the video signal field interpolation method of the present invention, the data set having the smallest difference between the maximum value and the minimum value is selected from among the plurality of data sets as an interpolation data set, which is suitable. The interpolation data set can be selected more easily.

In the video signal field interpolation method of the present invention, appropriate interpolation data can be obtained by using the average value of the interpolation data set as the interpolation data of the point to be interpolated.

According to the field interpolation method of the video signal of claim 5, two pixels existing at the same horizontal position on two vertically adjacent lines in the same field of a line including a point to be generated by interpolation. And at least a pair of pixels at positions that are respectively connected to the left and right of each pixel are extracted as reference pixels. If the reference pixel data in one of the two lines are similar to each other and the reference pixel data in the other line is monotonically increasing or monotonically decreasing, all points to be interpolated The average value of the data of two or more reference pixels in one of the two lines is used as the interpolation data. As a result, in the image data including the shallow slope boundary, misinterpretation of interpolation directions is reduced, and interpolation data with few errors can be obtained.

In the field interpolation method for a video signal according to a sixth aspect of the present invention, the data of reference pixels in one of the two lines are similar to each other, and the data of reference pixels in the other line are It is increasing monotonically,
After performing the above-mentioned process of making the interpolated data the average value of the data of two or more reference pixels in one of the two lines, the data of the reference pixel in the other line monotonically decreases. If it is, the image data that includes the thin slant line is determined and appropriate interpolation processing is performed. To be

[0019]

DESCRIPTION OF THE PREFERRED EMBODIMENTS A video printer to which the field interpolation method for video signals according to the present invention is applied, as shown in FIG.
RG that converts the video input signal of the to RGB analog signal
B converter 1, A / D converter 2 for converting RGB analog signals into RGB digital signals, and YMC converter 3 for converting RGB digital signals into YMC digital signals
A field memory 4 for storing a YMC digital signal for one field, an interpolator 5 for forming an image signal for one frame from the YMC digital signal for one field stored in the field memory 4, and a head 7 are controlled. And a head 7 for recording an image on a recording medium.

Next, the operation of the video printer will be described.

An NTSC video input signal is converted into an RGB analog signal by an RGB converter 1, and an RGB analog signal is converted into an RGB digital signal by an A / D converter 2. The RGB digital signal is converted into a YMC digital signal by the YMC converter 3, and the field memory 4 stores the YMC digital signal for one field. From the YMC digital signal for one field stored in the field memory 4, the interpolation unit 5 outputs 1
An image signal for a frame is formed, and the image is recorded on the recording medium by the head 7.

A first embodiment of the field interpolation method for video signals according to the present invention will be described below with reference to the drawings.

The pixels interpolated with the reference pixels are displayed as shown in FIGS. 2a and 2b. The matrix may be either a 5x2 matrix as shown in Figure 2a or a 3x2 matrix as shown in Figure 2b. The area example of each density data set to be used is as shown in FIGS. 3a and 3b, and FIGS. 3a (a) to (h) are the area examples in the case of 5 × 2 matrix reference, and FIG. (I) ~
Up to (p) are examples of regions in the case of 3 × 2 matrix reference. The image of the inclined boundary portion and the reference image data are as shown in FIG. 4, and the value X to be generated by interpolation is
And the density data D of the field image. An interpolation method for obtaining X _2,3 in the example of FIG. 4 when the density data set shown in the 5 × 2 matrix of FIG. 3A is used with the number of reference pixel data is 10 will be described.

The variance values of the concentration data sets (a) to (h) shown in the 5 × 2 matrix of FIG. 3a centering on the point X _2,3 are calculated for each set. The variance value is a value obtained by averaging the difference values between the respective data and the average value of the data, and for example, (a) D _1,3 , D _1,4 ,
When the variance values of D _1,5 and D _3,1 are calculated, average value = (D _1,3 + D _1,4 + D _1,5 + D _3,1 ) / 4 = (100 + 100 + 100 + 163) / 4 = 116 variance value = (| Average value -D _1,3 | + | Average value -D _1,4 | + |
Average value -D _1,5 | + | Average value -D _3,1 |) / 4 = (| 116-100 | + | 116-100 | + | 11
6-100 | + | 116-163 |) / 4 = 24. Similarly, when the sets (b) to (h) are calculated, the variance value of the a set = 24, the variance value of the b set = 31, the variance value of the c set = 38, the variance value of the d set = 50, e Variance value of set =
38, f set of variance values = 38, g set of variance values = 50, and h set of variance values = 47. Then, the density data set having the smallest density dispersion value is detected. Here, since the a group is the smallest, the a group is selected.

The average density value of the selected density data set is given at point X _2,3 .

X _2,3 = a mean value of a set = 116 As a result, the same effect as smoothing can be obtained, and interpolation data with a smooth slope with the jaggy angle reduced can be obtained, and misunderstanding in the interpolation direction can be reduced. As a result, interpolation data with little error can be obtained.

As described above, in this embodiment, since the work similar to the smoothing is performed at the time of the interpolation work, the same effect as the smoothing can be obtained. From the example showing the image of the inclined boundary and the frame data portion of FIG. 14,
The interpolation data portion shown in FIG. 5 can be obtained. Figure 5
The data obtained by this example shown in FIG.
The error is smaller than the interpolation data obtained by the conventional method shown in FIG. 5, and the accuracy is improved when the matrix of the field image data to be referred to is 3 × 2 or 5 × 2. Especially, the error in the case of 3 × 2 matrix is
The error is not so different from the error when the conventional 5 × 2 matrix is used. Therefore, in the case of the same matrix, the error is obviously suppressed.

Next, a second embodiment of the video signal field interpolation method of the present invention will be described with reference to the drawings.

The image of the inclined boundary portion and the reference image data are
Similar to the first embodiment, it is as shown in FIG. 4, and has a value X to be generated by interpolation and density data D of the field image. An interpolation method for obtaining X _2,3 in the example of FIG. 4 when the density data set shown in the 5 × 2 matrix of FIG. 3A is used with the number of reference pixel data is 10 will be described.

The difference value between the maximum value and the minimum value of each density data set (a) to (h) shown in the 5 × 2 matrix of FIG. 3a is calculated for each set. For example, D1, D
When 2, D3 and D4 are D1>D2>D3> D4, the difference value between the maximum value and the minimum value is calculated to be the difference value between the maximum value and the minimum value = D1-D4. Similarly, the difference value between the maximum value and the minimum value of each set is as follows.

Difference value between maximum value and minimum value of group a = 63 Difference value between maximum value and minimum value of group b = 100 Difference value between maximum value and minimum value of group c = 88 Maximum value and minimum value of group d Difference value of 100 = difference value of maximum value and minimum value of e set = 100 difference value of maximum value and minimum value of f set = 100 difference value of maximum value and minimum value of g set = 100 maximum value of h set Minimum value difference value = 100 The density data set having the smallest difference value between the maximum value and the minimum value is detected from each set. Since the density data set having the smallest difference value is the a set, the a set is selected.

The average density data of the selected density data set is given to the points X _2,3 .

X _2,3 = a mean value of a set = 116 As a result, the same effect as smoothing can be obtained, and interpolation data with a smooth slope with a jaggy angle reduced can be obtained, and misunderstanding in the interpolation direction can be reduced. As a result, interpolation data with little error can be obtained.

As described above, in this embodiment, the interpolation data set can be selected more easily than in the first embodiment. When interpolation data is obtained from the example showing the image of the inclined boundary and the frame data portion of FIG.
Similar to the first embodiment, the data shown in FIG. 5 is obtained.

Next, a third embodiment of the video signal field interpolation method of the present invention will be described with reference to the drawings.

The image of the inclined boundary portion and the reference image data are
As shown in FIG. 6, there are a value X to be generated by interpolation and density data D of the field image. A method of obtaining X _2,3 will be described with reference to the example of FIG.

As a first condition, it is checked whether or not the data located on the upper line or the lower line of the reference pixel data is an approximate value. In FIG. 6, the respective data on the upper line are similar.

D _1,1 ≈D _1,2 ≈D _1,3 ≈D _1,4 ≈D _1,5 Next, as the second condition, whether or not the data of the line on the upside down is monotonically increasing or not. Find out. In the case of FIG. 6, each data in the lower line monotonically increases.

D _3,1 <D _3,2 <D _3,3 <D _3,4 <D _3,5 and D _3,1- D _3,2 ≈ D _3,2- D _3,3 ≈ D _{3 , 3-} D _3,4 ≒ D _3,4 −
D _{3,5 When both} the first and second conditions are satisfied, the average value of the approximate data line is set as the interpolation value of the point X _{m, n} .

X _2,3 = (D _1,1 + D _1,2 + D _1,3 + D _1,4 +
D _1,5 ) / 5 = 100 When interpolation data is obtained from the data shown in FIG. 14 by the method described above, two points of data shown in FIG. 8 are obtained by this embodiment. Compared with the interpolation data according to the conventional example shown in FIG. 15, the error of the interpolation data of the image including the inclined boundary portion could be suppressed.

Next, a fourth embodiment of the video signal field interpolation method of the present invention will be described with reference to the drawings.

The image of the inclined boundary portion and the reference image data are
As shown in FIG. 7, there are points to be generated by interpolation and density data values of the field image. Here, an interpolation method in the case of obtaining X _2,3 after performing the above-described interpolation method of Embodiment 3 several times will be described.

As a first condition, it is checked whether or not the data located on the upper line or the lower line of the reference pixel data is an approximate value. In FIG. 7, each data on the upper line is similar.

D _1,1 ≈D _1,2 ≈D _1,3 ≈D _1,4 ≈D _1,5 Next, the second condition is whether or not the data on the line on the upside-down side monotonically decreases. Find out. In the case of FIG. 7, each data on the lower line monotonically decreases.

D _3,1 > D _3,2 > D _3,3 > D _3,4 > D _3,5 and D _3,1 −D _3,2 ≈D _3,2 −D _3,3 ≈D _{3 , 3-} D _3,4 ≒ D _3,4 −
D _{3,5 When both} the first and second conditions are satisfied, it is determined that the image data includes a slanted line, and in this case, the average value of D _1,5 and D _3,1 is set to the point X _{m, n.} And the interpolation value of.

X _2,3 = (D _1,5 + D _3,1 ) / 2 = 145 When the interpolation data is obtained from the data of FIG. 14 by the method described above, the interpolation data shown in FIG. 11 is obtained by this embodiment. To be Compared with the interpolation data according to the conventional example shown in FIG. 10, an error in the interpolation data of an image including a thin line could be suppressed. In addition, in FIG. 10A, | AF | and |
The interpolation result by the conventional interpolation method of extracting the average value of the extracted groups by extracting the group having the smallest difference between the three groups of B-E | and | C-D | FIG. 10B shows density data A to field pixel data.
The six values of F are binarized data, and the result and Fig. 1
By pattern matching using the pattern table of 3
The interpolation result by the conventional interpolation method of determining the direction to be interpolated and selecting either (A + F) / 2, (B + E) / 2, or (C + D) / 2 as the pixel density Y of the field is shown.

[0047]

In the field interpolation method of the video signal according to the first aspect of the present invention, the same effect as the smoothing can be obtained, and the interpolation data having a smooth slope with the jaggies can be obtained, and the misinterpretation of the interpolation direction can be reduced. As a result, interpolation data with little error can be obtained.

In the field interpolation method of the video signal of the second aspect, an appropriate interpolation data set can be selected.

In the video signal field interpolation method of the third aspect, an appropriate interpolation data set can be more easily selected.

In the video signal field interpolation method of the present invention, appropriate interpolation data can be obtained.

According to the field interpolation method of the video signal of claim 5, in the image data including the shallow slope boundary,
Interpolation data with less error can be obtained by reducing mistakes in the interpolation direction.

According to the field interpolation method of the video signal of the sixth aspect, in the image data including the thin inclined line, the misinterpretation of the interpolation direction is reduced, and the interpolation data with less error can be obtained.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of a video printer to which a video signal field interpolation method of the present invention is applied.

FIG. 2a is a diagram showing pixels interpolated with reference pixels of a 5 × 2 matrix.

FIG. 2b is a diagram showing pixels interpolated with reference pixels of a 3 × 2 matrix.

FIG. 3a is a diagram showing an example of a region of a data set of a 5 × 2 matrix according to the interpolation method of the present invention.

FIG. 3b is a diagram showing a region example of a data set of a 3 × 2 matrix according to the interpolation method of the present invention.

FIG. 4 is a diagram showing an example of data interpolated with reference pixel data of an image including an inclined boundary area.

FIG. 5 is a diagram showing field data and interpolation data of an image including an inclined boundary area according to the interpolation method of the present invention.

FIG. 6 is a diagram showing an example of data interpolated with reference pixel data of an image including a shallow slope boundary.

FIG. 7 is a diagram showing an example of data interpolated with reference pixel data of an image including a thin inclined line.

FIG. 8 is a diagram showing a part of field data and interpolation data of an image including a tilt boundary by the interpolation method of the present invention.

FIG. 9 is a diagram showing an example of field data of an image including a thin inclined line.

FIG. 10 is a diagram showing field data and interpolation data of an image including a thin slant line by a conventional interpolation method.

FIG. 11 is a diagram showing field data and interpolation data of an image including a thin inclined line according to the interpolation method of the present invention.

FIG. 12 is a diagram schematically showing a reference pixel and an interpolation pixel.

FIG. 13 is a diagram showing a conventional interpolation pattern table.

FIG. 14 is a diagram showing an example of field data of an image including an inclined boundary.

FIG. 15 is a diagram showing field data and interpolation data of an image including a tilt boundary by a conventional interpolation method.

FIG. 16 is a diagram showing jaggies caused by an error in density data of a conventional interpolation method.

[Explanation of symbols]

 4 Frame memory 5 Interpolator

Claims (6)

[Claims] 1. An interpolation method for forming an image signal for one frame from a digital image signal for one field stored in an image memory, in the same field of a line including a point to be generated by interpolation. Is an interpolation filter process for extracting, as reference pixels, two pixels existing at the same horizontal position on two lines adjacent to each other and at least a pair of pixels at positions contiguous to the left and right of each pixel, which is a reference of three or more points. A plurality of data sets each including data of two or more reference pixels including at least one data set including each pixel data are prepared, and one or more data sets based on the information of the plurality of data sets. Is selected as an interpolation data set, and the interpolation data of all points to be interpolated is determined from the interpolation data set. 2. The field interpolation method for a video signal according to claim 1, wherein a data set having the smallest data variance is selected from the plurality of data sets as an interpolation data set. 3. The field interpolation method for a video signal according to claim 1, wherein a data set having a smallest difference between a maximum value and a minimum value is selected from the plurality of data sets as an interpolation data set. 4. The field interpolation method for a video signal according to claim 1, wherein an average value of the interpolation data set is used as interpolation data of a point to be interpolated. 5. An interpolation method for forming an image signal for one frame from a digital image signal for one field stored in an image memory, in the same field of a line including a point to be generated by interpolation. Interpolation filter processing for extracting two pixels existing at the same horizontal position on two lines adjacent to each other and at least a pair of pixels at positions respectively connected to the left and right of each pixel as reference pixels, If the data of the reference pixels in either one of the lines are similar to each other and the data of the reference pixels in the other line are monotonically increasing or decreasing, the interpolated data of all the points to be interpolated are A field interpolation method of a video signal, which is an average value of data of two or more reference pixels in either one of the two lines. 6. The data of reference pixels in one of the two lines are approximate to each other, and the data of reference pixels in the other line monotonically increase. If the data of the reference pixels in the other line is monotonically decreasing after performing the above-mentioned processing to obtain the average value of the data of two or more reference pixels in any one of the two lines, The field interpolation method for a video signal according to claim 5, wherein the field interpolation method determines that the image data includes an inclined line and performs an appropriate interpolation process.