JPH06133280A - Scanning conversion device for picture signal - Google Patents

Abstract

PURPOSE:To improve the picture quality of a converted moving picture by applying motion compensation to an interpolation signal in response to a motion of a picture. CONSTITUTION:A motion vector detection circuit 25 detects a motion vector V0 from an input signal and the result is corrected by a correction circuit 26 and the vector a motion vector V used actually by interpolating scanning lines. The vector V consists of a horizontal motion vector Vx representing a motion component in the horizontal direction and a vertical motion vector Vy representing a motion component in the vertical direction. The input signal is also inputted to a motion compensation memory 7. The output from the memory 7 is a signal of a scanning line interpolated by motion compensation inter-field interpolation. Deterioration in picture quality in a moving picture is prevented by compensating the motion of a field delay according to the motion vector. Even when a picture has a motion, since scanning lines are interpolated without use of a vertical filter, the picture quality of moving picture conversion is excellent. The processing above is similar as to a motion component in the horizontal direction.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image signal scan conversion device, and more particularly to a device for converting a 2: 1 interlaced scan signal into a progressive scan signal.

[0002]

SUMMARY OF THE INVENTION The present invention is a device for scanning conversion of a television signal of, for example, 2: 1 interlaced scanning into a progressive scanning signal, in addition to a conventional inter-field or intra-field scanning line interpolation circuit, and motion compensation It has a motion-compensated interpolating circuit which has been applied, and improves the progressive scanning conversion image quality in a moving image by using these interpolating circuits by adapting to the motion of the image.

[0003]

2. Description of the Related Art Conventionally, this type of scanning conversion apparatus adapts the type of image when interpolating a scanning line required for conversion, and an inter-field interpolation circuit (static mode) is applied to a still image. ), An intra-field interpolator (dynamic mode) is used for moving images, and in some cases, an intermediate mode interpolator using a temporal-vertical spatiotemporal filter is used for slow-moving moving images.

Examples of conventional techniques include the following.

(1) Nishizawa, Tanaka, "Image quality improvement by progressive scanning conversion" December 1984, NHK STRL, (2) Achiha et al., "Motion adaptive signal processing for IDTV receivers", TV Academic journal, Vol. 41, No.
7 (1987) Moreover, there is the following as an example in which the motion compensation technique is used for the conversion of the number of frames of a television signal, though not the progressive scanning conversion.

(3) Tanaka et al., "Structure and image quality evaluation of HDTV-PASL conversion device using motion-compensated frame number conversion method", IEICE Transactions, vol. J
70-D, No. 8 (1987)

[0007]

The purpose of the progressive scan conversion apparatus is to convert an interlace signal into a progressive scan signal, and to disturb the image quality caused by the interlace scan (interlace disturb).
To improve the image quality. However, in the above-described conventional technique, as will be described later in detail, interlace interference is not removed from a moving image, and the degree of spatial improvement in the vertical direction is reduced to about 1/2. There was a drawback of not doing it. This situation does not make much difference even if the intermediate mode is used.

An object of the present invention is to improve this point and to provide a scanning conversion apparatus in which the motion compensation technology is utilized to improve the conversion image quality of a moving image.

[0009]

To achieve the above object, the present invention provides an intra-field interpolating circuit for interpolating a scanning line by inter-field signal processing for an input image signal of interlaced scanning, and an image in the input image signal. And a motion compensation interpolating circuit for interpolating a scanning line using a delay circuit that compensates the motion amount of the image according to the motion vector with respect to the delay amount of one field. , A mixing circuit that mixes the output signals of the two interpolation circuits, and a mixing ratio in the mixing circuit of 1
An interpolation selection determination circuit that determines based on a motion-compensated inter-frame difference signal of the input image signal in which the amount of motion of the image is compensated according to the motion vector with respect to the amount of frame delay,
It further comprises means for alternately selecting the output signal of the mixing circuit and the input image signal for each of the successive scans and outputting the successive scan signals.

[0010]

According to the present invention, the motion-compensated interpolated signal is mixed with the intra-field interpolated signal according to the state of motion detection of the image,
Used for progressive scan signals.

[0011]

Embodiments of the present invention will now be described in detail with reference to the drawings.

(First Embodiment) A. Description of FIG. 1 FIG. 1 shows the configuration of a progressive scan conversion apparatus according to a first embodiment of the present invention. The input signal of this device is a television image signal of 525 scanning lines, field frequency 60 Hz, 2: 1 interlace, and output signal is 525 lines, 60H.
z, a progressive scanning (interlace ratio 1: 1) television image signal.

The input signal is input to the motion vector detection circuit 25. The motion vector detection circuit 25 detects the amount of motion (motion vector) of the image from the input signal. The motion vector detection circuit 25 can be configured by a conventionally known technique, for example, the technique described in the document (3). The motion vector V ₀ detected by the motion vector detection circuit 25 is slightly modified by the motion vector value modification circuit 26, and becomes the motion vector V actually used in the interpolation of the scanning line. The operation content of the motion vector value correction circuit 26 will be described later.

The motion vector V actually consists of two signals, a horizontal motion vector V _x which represents the horizontal motion component of the image and a vertical motion vector V _y which represents the vertical motion component. The motion vector detection circuit 25 and the motion vector value correction circuit 26 constitute the motion vector generation circuit 6 as a whole. This apparatus can operate even if the value of the motion vector V is updated for each pixel, for each pixel block of a certain size, or for each field.

The input signal is also input to the motion compensation memory 7 having a delay amount of (263H + V). here,
"H" represents one line in the field. Further, “V” is a motion vector from the motion vector generation circuit 6. (26
The delay amount of (3H + V) is specifically expressed as follows.

(Condition) 1H = N pel (pel: number of pixels in pixel, N = 1 line) V _x : unit (pixel / field) V _y : unit ((number of scan lines in frame / field)) image with time Moves to the upper right, V _x > 0, V
_y > 0

[0017]

[Equation 1]

The output of the motion compensation memory 7 is the scanning line signal interpolated by the motion compensation inter-field interpolation. In this embodiment, the motion compensation memory 7 constitutes the motion compensation interpolation circuit 1 as a whole.

The input signal is also input to the 1H delay 8 and the adder 9. The adder 9 adds the output of the 1H delay 8 and the input signal of the device. The 1/2 coefficient unit 10 multiplies the output of the adder 9 by 1/2 and outputs the product. The 1H delay 8, the adder 9 and the 1/2 coefficient unit 10 constitute the intra-field interpolation circuit 2 as a whole. The output of the intra-field interpolation circuit 2 is the signal of the scanning line interpolated by the intra-field interpolation.

The output of the motion compensation interpolation circuit 1 and the output of the intra-field interpolation circuit 2 are input to a (1-K) coefficient unit 11 and a K coefficient unit 12, respectively, and multiplied by each coefficient. Where K
Is 0 ≦ K ≦ 1.0, and the value of K is input to these coefficient units 11 and 12 from an interpolation selection determination circuit 5 described later. The outputs of the (1-K) coefficient unit 11 and the K coefficient unit 12 are added by the adder 13, and finally become the signal of the interpolation scanning line.
These (1-K) coefficient unit 11, K coefficient unit 12, adder 1
3 constitutes the mixing circuit 3 as a whole.

The time axis conversion circuit 4 alternately selects the output of the mixing circuit 3 and the input signal of the present apparatus for each successive scanning to produce the sequential apparatus output signal of the present apparatus. 1H memories 14 to 17 and switches 18 to inside the time axis conversion circuit 4
The operation of 20 is clear from the related art, for example, Document (1), and thus detailed description thereof will be omitted. Briefly, for example, the first and second scanning lines are scanned with respect to the output sequential scanning signal. In time, adder 13
And the input signal of this device are written in the 1H memories 14 and 16, respectively. Then, at the time of the third and fourth lines, those signals are written in the 1H memories 15 and 17, the signals are read out from the 1H memory 14 as the third line, and the signals are selected as output signals by the switches 18 and 20.
Similarly, the signal read from the 1H memory 16 is used as an output signal for the fourth line.

The input signal is also input to the motion compensation memory 21 having a delay amount of (525H + 2V). The operation of the motion compensation memory 21 is such that the motion compensation amount is the motion vector V.
The delay amount is the same as that of the motion compensation memory 7, except that the delay amount can be easily inferred from the equation (1). The subtractor 22 is used for the input signal of the device and the motion compensation memory 2
The difference from the output of 1 is output. The output of the subtractor 22 is the motion compensation interframe difference signal for the input signal of the device.
The absolute value circuit 23 outputs the absolute value of the motion compensation inter-frame difference signal. The K determination circuit 24 determines the coefficient K in the coefficient units 11 and 12 based on the output of the absolute value circuit 23. The content of this K determination circuit 24 is simply the absolute value circuit 23.
The level conversion may be performed such that K is set to a large value if the output value of the above is large, or various methods of the conventional motion detection method as described in Document (2) may be incorporated. The motion compensation memory 21, the subtractor 22, the absolute value circuit 23, and the K determination circuit 24 collectively constitute the interpolation selection determination circuit 5.

B. Description of Operation Next, the operation of the apparatus of FIG. 1 will be described with reference to FIG. FIG. 2 is a view of the interlaced scan image at time (t) -vertical (y) coordinates, and the white circles in FIG. 2 indicate the positions of the scan lines generated by the interlaced scan. In the figure, the scanning of the image is performed from the top to the bottom in the same field (vertical line in the figure). In the figure, the vertical direction (y) is 0.
1- / 60sec when (sec) is the current field
Indicates 1 field before, and -1/30 sec indicates 2 field before. Now, focusing on the scanning line a in FIG. 3A, in order to convert this image into a progressively scanned image, it is necessary to create an interpolation scanning line x in the middle of the scanning line b immediately above in interlaced scanning. There is.

In the conventional progressive scan conversion, when the image is stationary, for example, the scanning line of d one field before is set to 1
Field delay is performed to form an interpolated scanning line x (inter-field interpolation).

[0025]

[Outer 1]

Was used as the interpolating scanning line x (in-field interpolating). However, in the intra-field interpolation, since the interpolation scan line x is created by the vertical low pass filter,
The pass band of the vertical spatial frequency is reduced to 1/2, and the vertical resolution of the image is deteriorated. At the same time, the interlace interference is not removed as will be described later. Therefore, although the conventional device has good image quality for still images, the image quality of moving images deteriorates.

In the apparatus of FIG. 1, the image quality deterioration in a moving image is solved by motion compensation of the field delay amount according to the motion vector V. For example, when the vertical motion of the image is V _y = 2, the motion compensation memory 7 of FIG. 1 causes the scanning line c in FIG. 2A (c is the number of scanning lines in the frame with respect to x). 2) is delayed to form the interpolated scanning line x. As described above, the apparatus of FIG. 1 can interpolate the scanning lines without using a vertical filter even when there is a motion in the image, so that the converted image quality of the moving image is good. This situation also applies to the horizontal motion component. Since V _x = V _y = 0 in a still image, inter-field interpolation for a conventional still image is also included in the above motion compensation interpolation operation.

By the way, FIG. 2A shows only when V _y is an even number, but when V _y is an odd number, there is no scanning line at a position to be interpolated one field before, and therefore motion compensation interpolation is performed. I can't do it. Therefore, the motion vector value modification circuit 26 of FIG. 1 modifies the motion vector amount in the vertical direction by the following algorithm. When the vertical and horizontal motion vectors detected by the motion vector detection circuit 25 are V _y0 and V _x0 , respectively,

[0029]

[Equation 2]

[0030]

Equation 3] _{V x = V x0 ... (3} ) i.e., for V _y0 are if V _y0 is odd are rounded to a small even number of more absolute value V _y0. This is because the image quality is generally better in the under-compensated state than in the over-compensated state when the motion compensation is not accurate. With respect to the horizontal motion component, it is not necessary to modify V _x0 because there is no special situation based on the interlaced scanning as described above.

The detected motion vector is not always accurate, and even if it is accurate, the value may be corrected as described above. Furthermore, there may be a plurality of motions in the same image portion where the detected motion vector is constant. In such a case, the motion compensation interpolation may cause a large image quality deterioration. Therefore, even when the motion compensation interpolation is used, a large image quality deterioration can be eliminated by using the interpolation by the intra-field interpolation circuit 2 together. At this time, the amount of each of the interpolation by the motion compensation interpolation circuit 1 and the interpolation by the intra-field interpolation circuit 2 is determined by the output K of the K determination circuit 24, and K is the output of the subtractor 22 which is motion compensation. It is determined according to the inter-frame difference signal. That is, if the motion compensation by the motion vector works well, K will be close to zero, and as a result,
If motion compensation interpolation is superior to intra-field interpolation and conversely motion vector compensation does not work well, K
Becomes close to 1, and as a result, intra-field interpolation is superior to motion compensation interpolation. FIG. 2B shows this situation. When the signal values at points a, b ... In FIG. 2B are expressed as Pa, Pb ..., the motion vector value is V _y.
= 2, i.e. in the case of _{2V y = 4 | Pa-Pe} | K based on the magnitude is determined. When V _y = 0, K is |
Pa-Pf | determined by further if V _x = 0, the operation of the interpolation selection determination circuit 5 is equivalent to the operation of the motion detection circuit in a conventional device.

FIG. 3 shows the time-vertical spatiotemporal spectrum of the operation of the conventional device and the device of FIG.
The horizontal axes of (a) to (e) in FIG. 3 are all temporal frequencies f (unit: Hz), and the vertical axes are vertical spatial frequencies ν (unit: c).
ph, cycles perpicure heig
ht).

FIG. 3 (a) shows the spectra of the images considered below. The same figure

[0034]

[Outside 2]

Represents an image moving in the vertical direction at a speed of V _y = 2. The original spectrum of the image extends linearly from the origin to the upper left, and the folding spectrum of the sampling in the time direction by the field also extends to the upper left from the point (f, ν) = (60, 0).

[0036]

[Outside 3]

Is a carrier for spatiotemporal sampling by interlaced scanning, and a folding component due to it extends from each x to the lower right. The spectrum of the aliasing due to the interlace is shown by a dotted line in the figure, and this is the component that causes the image quality disturbance due to the interlace. The interlace-sequential scanning conversion is for removing the spectrum of the dotted line by scanning conversion to improve the image quality of the image.

FIG. 3B shows a frequency response of inter-field interpolation for a conventional still image. The shaded area in the figure is the stopband when the interpolation is viewed as a filter.
From this figure, it can be seen that in this interpolation method, part of the original spectrum is cut (the image is blurred), and conversely, part of the dotted spectrum remains (interlace interference is not removed). FIG. 3C shows the frequency response of the conventional intra-field interpolation. Although the stop band is different, the image remains blurred and the interlace interference remains. In some cases, the conventional apparatus uses interpolating modes between fields and intermediate fields, as in the examples of documents (1) and (2). However, the stop band in this case is generally represented by a triangular region as shown in FIG. 3D, but it can be seen that there is no great difference in the results from those in FIGS. 3B and 3C.

On the other hand, the frequency response of the motion compensation interpolation used in the apparatus of FIG. 1 is as shown in (e) of FIG. From the figure, the original image spectrum of the solid line is not damaged at all. It can be seen that the folded spectrum of the dotted line is completely removed. That is, since the original image component is reproduced in full resolution and the interlace interference is completely removed, the conversion image quality in motion can be improved significantly.
In FIG. 2E, an image of V _y = 2 is shown, but V _y = 2
If n (n is an integer), the image quality can be similarly improved. Regarding horizontal movement, the image quality can be similarly improved regardless of any integer value of V _x . Of course, even if the movement of the image is in the oblique direction (V _x ≠ 0, V _y ≠ 0), the image quality can be similarly improved if V _y = 2n.

(Second Embodiment) FIG. 4 shows a motion compensation interpolation circuit according to a second embodiment of the present invention. To implement this embodiment, for example, the motion compensation interpolation circuit 1 in FIG. 1 is replaced by the circuit in FIG. 4, and the motion vector value correction circuit 26 in FIG.
May be removed and the output V ₀ of the motion vector detection circuit 25 may be directly used as the output of the motion vector generation circuit 6 (V _y = V
_{y0, V x = V x0)} . Figure 1 of the motion compensated interpolation circuit 1 V _y
If the number is odd, interpolation could not be performed.
The circuit of (1) improves this point.

In FIG. 4, the input signal is a motion compensation memory 31 having a delay amount of (263H + V ') and (263H + V).
V ″) is input to the motion compensation memory 33. In the memory 31, the vector value from the motion vector generation circuit 6 used for motion compensation has a horizontal component of V _x = V _x.
However, the vertical component V _y ′ is provided as the output of switch 32. The switch 32 outputs V _y from the same circuit 6.
According value, if mod ₂ V _y (remainder of the V _y divided by 2) is 0 outputs V _y, and outputs the V _y +1 if 1. On the other hand, in the memory 33, V _x ″ = V _x , V
The motion compensation of _y ″ = V _y −1 is performed. The output of the memory 31 is guided to the switch 36 and the adder 34. The adder 3
In 4, the output of the memory 31 and the output of the memory 33 are added. The subsequent 1/2 coefficient unit 35 outputs 1/2 to the output of the adder 34.
Is multiplied to obtain an input on one side of the switch 36. Switch 3
6 selects the output of the memory 31 when mod ₂ V _y is 0, and selects the output of the 1/2 coefficient unit 35 when 1 is mod ₂ V _y . The output of the switch 36 becomes the motion compensation signal. Note that V _y −
1, V _y +1 and mod ₂ V _y can be provided in the motion vector generation circuit 6 by providing a circuit for generating them based on the output of the motion vector detection circuit 25.

As is clear from the above operation, the motion compensation interpolation circuit of FIG. 4 is equivalent to the motion compensation interpolation circuit 1 of FIG. 1 if V _y is an even number. When V _y is an odd number, the memory 31, the memory 33, the adder 34, and the 1/2 coefficient unit 35 form a space-time filter, and the output of the filter is used as an interpolation signal. Although FIG. 4 shows the case of the second order as this filter, it is of course possible to configure a filter with a high order. In that case, a large number of motion compensation memories may be provided and the motion vectors of V _y ± 1, V _y ± 3, V _y ± 5 ...

FIG. 5 illustrates the operation of the circuit of FIG. 4 when V _y = 1. The notation of FIG. 5 is the same as that of FIG. Since V _y = 1, V _y ′ = V _y + 1 = 2
V _y ″ = V _y −1 = 0. Therefore, the motion compensation memory 3
1 performs the operation of moving the scanning line c in FIG. 5 to the interpolation point x, and the motion compensation memory 33 performs the operation of moving the scanning line d to x. Interpolation scan line

[0044]

[Outside 4]

Is given by At this time, the motion compensation inter-frame difference signal required for the interpolation selection determination is Pg-Pa.
However, since the movement between a and g is 2V _y = 2 (even number), the operation of the interpolation selection determination circuit 5 in FIG. 1 can be performed without any particular problem. That is, the circuit 5 can operate normally regardless of any integer value of V _y .

FIG. 6 also shows the spatiotemporal spectrum when V _y = 1. When V _y is an odd number, the original spectrum and the folded spectrum completely overlap each other on the f−ν region, and therefore cannot be separated by any interpolation filter. 6A shows the frequency response of the conventional intra-field interpolation, and FIG. 6B shows the frequency response of the motion compensation circuit of FIG. In both figures, the suppressed component and the remaining component are equal. Therefore, the converted image quality is also the same. However, the actual image often has complicated movements, and V _y
It is also conceivable that the movement of = 1 and the movement of V _y are mixed. In such a case, as is clear from FIG. 6 (b), the circuit of FIG. 4 can detect the vector of V _y = 1 even if it is detected by the motion vector detection circuit and used for motion compensation. It has an advantage that a visually noticeable disturbance near (f, ν) = (± 30, 0) caused by an even number of movements of V _y can be removed.

(Third Embodiment) FIG. 7 shows the arrangement of a progressive scan conversion apparatus according to the third embodiment of the present invention. The image input signal and output signal of this device are the same as in FIG. In this apparatus, the amount of motion (motion vector) of the image of the image input signal is further given as a motion vector input signal. In this case, it is assumed that motion vector detection is performed on the image transmission side of the image, and the motion vector is transmitted together with the image signal. As an example in which such an assumption holds, for example, in digitally encoded transmission of an image, which has been attracting attention in recent years, a motion vector is transmitted as a part of a transmission code.

The motion vector input signal is format-converted in the motion vector receiving circuit 79 into a format required for this apparatus, and the value is corrected in the motion vector value correcting circuit 80,
The motion vector V (V _x , V _y ) used in this device is distributed to each part of the device. These circuits 79 and 80 together constitute a motion vector generation circuit 78. The operation content of the motion vector value correction circuit 80 is the same as that of the motion vector value correction circuit 26 of FIG. This circuit 78 is used in place of the motion vector generation circuit 26 of the embodiment of FIG.

The image input signal is input to the cascade connection portions of fixed field delays 41 to 43 (however, 41 and 43 are 262H delays and 42 is 263H delays). Image input signal and output of field delay 42

[0050]

[Outside 5]

Is input to. These operations are similar to those of the motion compensation memory 7 of FIG. However, since the standard delay amount is 16H, which is smaller than that of the motion compensation memory 7, it is economically advantageous in realizing an expensive motion compensation memory. The value of 16H is determined in consideration of the fact that the moving speed V in the normal case rarely exceeds 16 (H / field). The outputs of the motion compensation memories 46 and 47 are added by the adder 48. The 1/2 coefficient unit 49 multiplies the output of the adder 48 by 1/2. The memories 46 and 47, the adder 48, and the 1/2 coefficient unit 49 constitute a motion compensation interpolation circuit 50 as a whole.

The image input signal and the output of the field delay 42 are also input to the 16H memories 52 and 51, respectively. Memories 51, 52, adder 53, 1/2 coefficient unit 54
The operation of is similar to that of the motion compensating interpolation circuit 50 except that no motion compensation is performed, and these constitute the inter-field interpolating circuit 55 for a still picture as a whole.

The output of the field delay 41 is input to the 16H memory 56. Memory 56 and said memory 5
Reference numerals 1 and 52 are all for delay adjustment with the motion compensation memories 46 and 47. The output of the 16H memory 56 is input to the intra-field interpolation circuit 57. The contents of the intra-field interpolation circuit 57 may be the same as the contents of the intra-field interpolation circuit 2 of FIG. 1, for example.

The output of the motion compensation interpolating circuit 50 and the output of the inter-field interpolating circuit 55 are input to the 1 side of the switch 58,
It becomes 0 side input. The switch 58 selects the 1 side when the signal L is 1, and the 0 side when the signal L is 0 according to the signal L described later. The output of the switch 58 becomes the input of the (1-K) coefficient unit 59, and the output of the intra-field interpolation circuit 57 becomes the input of the K coefficient unit 60. 1-K coefficient unit 59, K coefficient unit 6
The operation of 0 and the adder 61 is similar to that of the mixing circuit 3 in FIG. 1, and a switch 58 is added to these to form a mixing circuit 62 as a whole. Output of adder 61 and 16H
The output of the memory 56 is input to the time axis conversion circuit 63.
The contents of the time axis conversion circuit 63 may be the same as the contents of the time axis conversion circuit 4 of FIG. 1, for example. The output of this circuit 63 becomes the output signal of this device.

In FIG. 7, motion compensation memories 46 and 4
The outputs of the 7 and 16H memories 51, 52 and 56 are respectively labeled with C ', A', C, A and B, which are the corresponding inputs of the subtractors 65 to 68, respectively. . On the other hand, the output of the 262H memory 43 is (16H + 2
It is input to the motion compensation memory 44 and the 16H memory 45 having the delay amount V). The outputs of these memories 44 and 45 are labeled D and D ', but they are also the corresponding inputs of the subtractors 66 and 68. Subtractor 6
Outputs 5 and 66 are inter-frame difference signals of the input image signal (without motion compensation), and outputs of subtracters 67 and 68 are motion-compensated inter-frame difference signals of the input image signal.

The outputs of the subtracters 65 and 66 and the subtracters 67 and 68 are out of phase with each other by one field. The method of using the inter-frame difference signal with such an out of phase for the determination is in motion detection. It has been conventionally used to prevent detection failure (for example, Document (3)). Subtractors 65-6
The absolute values of the eight outputs are obtained by the absolute value circuits 69 to 72, respectively. The maximum value circuit 73 compares the output of the absolute value circuit 69 with the output of the absolute value circuit 70 and outputs the one with the larger value. The operation of the maximum value circuit 74 is similar, and the outputs of the absolute value circuits 71 and 72 are compared.

The L determining circuit 75 compares the output of the maximum value circuit 73 with the output of the maximum value circuit 74 to determine the constant L.
For example, if the output of the circuit 73 is larger than the output of the circuit 74, L = 1, and vice versa. Here, L is a binary value of 0 or 1, but in the mixing circuit 62, 1-K coefficient unit 59 and K coefficient unit 6 are used instead of the switch 58.
0, if a circuit similar to the adder 61 is provided, L can be a continuous value.

On the other hand, the output of the maximum value circuit 73 and the output of the maximum value circuit 74 are also input to the minimum value circuit 76. The minimum value circuit 76 outputs the smaller of the two inputs. The output of the minimum value circuit 76 is input to the K determination circuit 77. K
The operation of the decision circuit 77 is similar to that of the K decision circuit 24 of FIG. The output of the K determination circuit 77 becomes the coefficient K input to the mixing circuit. Each of these components 65 to 77 constitutes the interpolation selection determination circuit 64 as a whole.

Next, the operation of the apparatus shown in FIG. 7 will be described with reference to FIG. FIG. 8 shows the positions of the signals indicated by A, A ′, ... In FIG. 7 in the space-time domain. The scan line of interest is B, and the interpolation scan line for it is X.
Other notations are the same as in FIG. In the apparatus of FIG. 7, both inter-field interpolation and motion compensation inter-field interpolation are performed from both sides in the time direction (the apparatus of FIG. 1 is one-sided interpolation). That is, in inter-field interpolation

[0060]

[Outside 6]

In the motion compensation inter-field interpolation,

[0062]

[Outside 7]

[0063] The intra-field interpolation is the same as in FIG. By performing the two-sided interpolation with respect to time, the present apparatus can obtain a steeper attenuation characteristic with respect to time frequency. However, the stop band due to each interpolation on the spatial frequency domain is shown in FIG.
(B), (c), and (e) are completely the same.

On the other hand, in the present apparatus, the interpolation selection determination circuit 64 determines which of the three interpolations is to be selected. This is done by the inter-frame difference signals AC, BD and the motion compensation frame. 4 of inter-signals A'-C 'and BD'
It is based on one signal. These spatiotemporal phases are also clear from FIG. From these signals L, K
However, the method and meaning thereof will be omitted because they are clear from the description relating to FIGS.

In the present apparatus, inter-field interpolation for a still picture is provided separately from motion compensation interpolation, so that a still picture part is generated in the same picture part where the motion vector V output from the motion vector generating circuit 78 is constant. Even when the moving image portion and the moving image portion are present at the same time, they can be scan-converted with high image quality. Note that the motion compensation interpolation circuit of FIG. 4 can be applied to this device in the same manner as in the device of FIG.

The present invention can be applied not only to the image signals shown here, but also to, for example, a 625-line system or HDTV. Further, although only one motion compensation interpolation circuit is used in each of the above-mentioned embodiments, in order to cope with more complicated motion of an image, the motion compensation interpolation circuit is processed by the same method as that of the embodiment of FIG. 4 or 7. The number of can be increased. For example, in FIG. 7, if a motion compensation memory is used instead of the 16H memories 45, 51 and 52 and the motion vector generation circuit 78 is changed so that two motion vectors can be output at the same time, two types can be obtained for the same image portion. It is possible to perform progressive scan conversion with good image quality up to the movement.

[0067]

As described above, according to the present invention, when interpolation of scanning lines necessary for converting an interlaced scanning signal into a progressive scanning signal is performed, motion compensation is performed by adapting the motion of an image to the interpolation signal to produce a moving image. On the other hand, the interlace interference can be completely removed without lowering the vertical resolution, and the converted image quality of the moving image can be greatly improved.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a diagram showing an operation example of the embodiment.

FIG. 3 is a diagram showing a time-vertical spectrum for explaining the operation of the embodiment.

FIG. 4 is a diagram showing a motion compensation interpolation circuit according to another embodiment of the present invention.

FIG. 5 is a diagram showing an operation example of the embodiment.

FIG. 6 is a diagram showing a time-vertical spectrum for explaining the operation of the circuit of FIG. 4;

FIG. 7 is a view showing still another embodiment of the present invention.

FIG. 8 is a diagram showing an operation example of the embodiment.

[Explanation of symbols]

 1 motion compensation interpolation circuit 2 intra-field interpolation circuit 3 mixing circuit 4 time axis conversion circuit 5 interpolation selection determination circuit 6 motion vector generation circuit

Claims (4)

[Claims] 1. An inter-field interpolation circuit for interpolating a scanning line by intra-field signal processing for an input image signal of interlaced scanning, and a motion vector generation circuit for outputting a motion amount of an image in the input image signal as a motion vector. A motion compensation interpolation circuit that interpolates a scanning line using a delay circuit that compensates the motion amount of the image according to the motion vector with respect to the delay amount of one field; and a mixing that mixes output signals of the two interpolation circuits. Circuit, and an interpolation selection determining circuit for determining a mixing ratio in the mixing circuit based on a motion-compensated inter-frame difference signal of the input image signal in which a motion amount of the image is compensated for a delay amount of one frame according to the motion vector. And means for alternately selecting the output signal of the mixing circuit and the input image signal for each of the successive scans and outputting the successive scan signals. Scan conversion apparatus for an image signal, characterized in that comprises a. 2. The motion vector generation circuit has an odd amount of motion of the motion vector in the vertical direction (scanning line / field).
2. The scan conversion device for an image signal according to claim 1, wherein a value obtained by adding the motion amounts -1 and +1 is output as a vertical motion vector. 3. The motion compensation interpolation circuit, when the motion amount of the motion vector in the vertical direction is an even number (scanning line / field), the motion compensation interpolating circuit calculates the delay amount from the signal one field before or after the delay amount by the motion amount. An output signal is generated, and in the case of an odd number (scanning line / field), the output signal is generated from a signal one field before or after the delay amount is compensated by an amount obtained by adding or subtracting an odd amount to the motion amount. Claim 1
An image signal scanning conversion device as set forth in. 4. An inter-field interpolating circuit for interpolating a scanning line by each signal before and after one field of the input image signal, wherein a delay circuit in the motion compensation interpolating circuit is provided before and after one field of the input image signal. The amount of motion of the image is compensated for each signal according to the motion vector, and the mixing circuit selects either the output signal of the motion compensation interpolation circuit or the output signal of the inter-field interpolation circuit by a selection signal. The interpolation selection determination circuit mixes with the output signal of the intra-field interpolation circuit, and the interpolation selection determination circuit compensates the motion amount of the image with respect to the inter-frame difference signal of the input image signal and the delay amount of one frame. 2. The image signal according to claim 1, wherein the selection means is controlled and the mixing ratio is determined based on a compensation frame difference signal. Scan converter.