JPS6330079A - Picture signal converter - Google Patents

Abstract

PURPOSE:To contrive to prevent fog in following to a duplicated image in following to a moving object picture by obtaining a difference between a noted field and its preceding field and a difference between the noticed field and its succeeding field and switching an inter-field interpolation signal properly based on th result of movement discrimination. CONSTITUTION:If movement exists in the field (i+1), since the correlation with th field (i) is weak, the absolute of signals (x1-x2) and (x1-x3) is larger than the threshold value Th, difference discrimination signals c, d become H and then an interpolation selection signal S2 becomes H. When both the fields (i-1) and (i) are in standstill, the correlation between them is strong, and since either of the signal (x4-x2) or (x4-x3) is smaller than the value Th, the interpolation selection signal S1 becomes L. Then a selector 50 outputs a signal x4 as an interpolation signal x7. In case the field (i-1) is moved and the fields (i) and (i+1) are in standstill, the signal S1 becomes H and the signal S2 becomes L and the selector 50 outputs the signal x1. In case the field (i) is moved, since the correlativity between the fields (i-1) and (i+1) is weak, all discrimination signals a d are at H, the signals S1, S2 become H and the selector 50 outputs a value (x2+x3)/2.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an apparatus for converting an image signal such as a television signal from inclace scanning to non-interlaced scanning.

[Conventional technology and its problems]

As a method for obtaining a high-definition image from a current interlaced video signal such as NTSC, a non-interlaced method has been proposed in which the scanning lines of one field are doubled, for example.

In order to double the number of scanning lines, the problem is how to generate an interpolation signal between the scanning lines. A method has been proposed in which interfield interpolation uses signals corresponding to fields as they are as interpolation signals, and switches between them in accordance with the movement of an image. In this method, the accuracy of motion detection of the object image is an important factor that affects the image quality, but in -S, the presence or absence of motion is determined by whether or not there is a significant difference in the difference signal (frame difference) between the front and rear fields. Judging. However, in the current field, a significant difference may not appear in the difference signal between the preceding and succeeding fields even in a motion region, and in such a case, such a determination method may cause significant image quality deterioration such as double images.

This image quality deterioration will be explained with reference to FIGS. 5 and 6.

(1) to (6) in FIG. 5 correspond to (1) to (6) in FIG. 6. In FIGS. 5 and 6, the shaded areas indicate selection of intra-field interpolation. Now, in field i-1 (2)
The object image shown by the thick line, which was stationary at the position, moves from ■ to ■ in field i, and changes to (4) in field i+1.
). At this time, if motion detection is performed using the difference signal between the front and rear fields, in the area shown by ■ and ■ in field i in FIG.
, Intra-field interpolation is selected because there is a significant difference in the frame differences of ■, but since there is no significant difference in the frame differences of In the area marked ■ in FIG. 5, a signal corresponding to the background image of the previous field is interpolated, causing double image interference.

Also, movement occurs only in field i,
Since field i-1 and field i+1 are stationary, inter-field interpolation is possible, but motion detection using frame difference cannot determine the field in which motion is occurring, so for these fields i-1 and field i+1, interpolation is possible. This method has the disadvantage that it involves intra-field interpolation, resulting in an unnatural image in which low-resolution images (blurred images) follow the front and back of the moving object.

SUMMARY OF THE INVENTION An object of the present invention is to overcome the above-mentioned drawbacks and provide an image signal conversion device capable of obtaining high-definition non-interlaced images without double image interference or blurred images following moving objects. .

[Means for solving problems]

An image signal conversion device according to the present invention includes: a first forming means for forming an interpolated signal from a signal of the same field; a second forming means for forming an interpolated signal from a signal of a previous field;
a third forming means for forming an interpolated signal from a signal of a subsequent field; a fourth forming means for forming an interpolated signal from signals of a previous field and a subsequent field; and a previous field preceding a field of interest for which an interpolated signal is to be formed. determining means for determining the motion of the image by calculating a difference signal between adjacent fields between the signal of the field of interest, the signal of a subsequent field following the field of interest, and the signals spatially located above and below the interpolation scanning line; Based on the determination result, one of the interpolation signals is appropriately selected.

[Effect]

The double image disturbance occurs because the movement of the image is detected by frame differences, and therefore the movement of the field between them cannot be detected in principle.

In addition, in the blurred image, it is not possible to specify which field of the two fields in which the frame difference is calculated, in which motion has occurred, and the signal of the previous field is uniformly used as the interfield interpolation signal. This is caused by Therefore, the present invention focuses on field differences, and uses the strong inter-field correlation in the case of still images to remove the double image disturbance, and also removes the tracking of the blurred image by using the field of interest. In addition to the difference between the field of interest and the previous field, the difference between the field of interest and the subsequent field is also determined, and based on the movement judgment results, the interfield interpolation signal is switched appropriately to track double images and moving object images. Blur can be effectively prevented.

〔Example〕

Hereinafter, the present invention will be described using one embodiment of the present invention with reference to the drawings. In this example, the input image signal is
It is assumed that one frame is composed of 525 lines.

FIG. 1 is a block diagram of one embodiment of the present invention. The image signal x1 input to the input terminal 10 is sent to the field memory 12. Line memory 14. field memory 16
are delayed by 262 lines, 1 line, and 262 lines, respectively. The output of field memory 12 is
, the output of the line memory 14 is X=, and the output of the field memory 16 is x4.

The subtracter 18 subtracts the signal x2 from the image signal x4 263 lines before it, and the absolute value circuit 20 takes the absolute value of the output of the subtracter 18 and supplies it to the comparison circuit 22.

The subtracter 24 subtracts the image signal x3 from the image signal x4 one field (actually 262 lines) before it, and the absolute value circuit 26 takes the absolute value of the output of the subtracter 24 and supplies it to the comparison circuit 28. do. Comparing circuits 22 and 28 compare their inputs with a threshold Th, and output difference determination signals b and a (H when greater than Th, L when less than Th). A N
The D circuit 30 calculates the AND of the outputs of the comparator circuits 22 and 28, and outputs an interpolation selection signal SI.

The subtracter 32 subtracts the image signal X from the image signal x2 one field (actually 262 lines) before it, and the absolute value circuit 34 takes the absolute value of the output of the subtracter 32 and supplies it to the comparison circuit 36. do. The subtracter 38 extracts the image signal X and the image signal X one field (actually 263 lines) before it.
, and the absolute value circuit 40 takes the absolute value of the output of the subtracter 38 and supplies it to the comparison circuit 42. Comparison circuit 3
6 and 42 compare the input with the threshold Th, and generate a difference judgment signal (H when it is greater than Th, L when it is less than Th) c,
Output d. It should be noted that the comparator circuits 36 and 42 have opposite input sides for the dark value Th. A N
The D circuit 44 calculates the AND of the outputs c and d of the comparison circuit 36.42, and outputs an interpolation selection signal S2.

On the other hand, the adder 46 adds x4 to the video signal
The /2 coefficient circuit 48 multiplies the sum by 172 and supplies the interfield interpolation signal (average value of the previous and subsequent fields) Xs to the selector 50. The adder 52 adds the video signal x1 to the video signal x2, the 1/2 coefficient circuit 54 multiplies the sum by 172, and the intra-field interpolation signal X is sent to the selector 50.
supply to. The selector 50 is also supplied with the input video signal Xt and a video signal x4 delayed by one frame (525 lines), and interpolation selection signals St and Sz are applied as control signals.

The selector 50 receives these interpolation selection signals S, .

One of the four interpolation signals is selected according to S2 and supplied to the time compression circuit 56. Interpolation selection signal S.

, St and the selected interpolation signal x7 are shown in FIG. The time compression circuit 56 compresses the time axes of the original signal x3 and the interpolation signal x7 from the selector 50 to 1/2, and sequentially and alternately outputs the compressed signals to the output terminal 58.

Next, each difference determination signal a between consecutive fields,
b, c, d and video signal x, + xz l
The operation of the circuit shown in FIG. 1 will be explained with reference to FIG. 2 which shows the relationship with x31x4. Furthermore, the difference determination signal a in FIG.
, b, c, d and video signal xl + xZ + X3
.

x4 corresponds to the difference determination signals a, b, c, d and the video signal xI+xZ+X:llXa in FIG. 1, respectively. In Figure 2, the vertical axis indicates the vertical direction on the screen,
The horizontal axis shows time. Therefore, vertically arranged signals are
in the same field. Here, a description will be given of which interpolation signal is used to interpolate between the video signal x2 and the video signal X of field i.

Suppose now that the field is i+1 forceful movement. Since there is almost no correlation between field i and field i+l, the absolute value of (xt xz) is greater than Th, the difference determination signal C becomes H, and (xt −X3
) is also larger than Th, and the difference judgment signal d is also “H”.
”, and as a result, the interpolation selection signal S2 becomes “H”. If both fields i-1 and i are “stationary”,
If there is a strong correlation between field i-1 and field i, then x
Even if an image edge exists between t and x3, (Xa
Since either the absolute value of xz) or the absolute value of (Xa X3) is considered to be smaller than Th, the interpolation selection signal Sl becomes L. Therefore, as shown in FIG. 7, the selector 50 outputs the signal x4 as the interpolation signal x7.

Conversely, when field t-1 moves by one movement, fields t and i
If +1 is "stationary", Sl will be H, S2 will be L,
The selector 50 outputs the signal x1 as the interpolation signal x7.

If fields i-1, i, and i+1 are all "stationary 1",
Since the interfield correlation is strong, the interpolation selection signals S+, Sz
Both become L, and the selector 50 outputs the average value (xi +X4)/2 between the preceding and following fields as the interpolation signal X7.

If field i is "movement", between field i-1 and field i, and between field i and field i+l
Since there is a weak correlation between the judgment signals a, b,
c, d are all H, and interpolation selection signals St, Sz
Both become H, and the selector 50 outputs (xz +X3)/2 between the inner fields as the interpolation signal X.

In this embodiment, FIG. 3 shows how an image is displayed on the screen, and FIG. 4 shows how interpolation is performed in consecutive fields. In Figure 3, the shaded area indicates the intra-field interpolation area, the area marked as front F interpolation indicates that the interpolated signal is taken from the previous field, and the part marked as post F interpolation indicates that the interpolated signal is taken from the next field. show. Fourth
In the figure, O is a normal scanning line signal, MU indicates interpolation using an intra-field signal, ・ indicates interpolation using the average value of the signals of the previous and previous fields, K indicates interpolation using the previous field signal, ◆ indicates interpolation using the signal of the subsequent field. Further, the arrows in FIG. 4 indicate which scanning line signals are used as interpolation signals.

As can be seen from FIG. 3, in this embodiment, the inner field area is selected only for field i where movement has occurred, and double images as explained in FIG. 5 do not occur. do not have. Also, as can be seen from Figure 4,
If no movement occurs in three consecutive fields, the center field is interpolated using the average of the preceding and following fields. Since the interpolation using the average value improves the S/N ratio of the signal, the S/N ratio of the entire image is also improved. Since most of the image signal is usually a static area, interpolation using the average value of the preceding and following fields is extremely effective.

〔Effect of the invention〕

As can be easily understood from the above explanation, the present invention removes images with low resolution that follow moving object images,
Furthermore, scanning line conversion without image edge deterioration or double images can be realized.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention, and FIG. 2 is a block diagram of an embodiment of the present invention.
Field difference determination signals a, b, and c obtained by the circuit of FIG. 1 for three consecutive fields. d and video signal x, + xz l xi l xi
FIG. FIG. 3 11>≠n is a diagram showing what kind of interpolation is performed on the screen due to the movement of an object according to the present invention, and FIG. FIG. 3 is a diagram showing what kind of interpolation is adopted. 5 and 6 are diagrams for explaining the interpolation signal selection method and its effects by the conventional device, and FIG. 7 shows the relationship between the interpolation selection signals S+ and Sz in FIG. 1 and the selected interpolation signal X. FIG.

Claims (1)

[Claims] An image signal conversion device that inserts an interpolation signal into an image signal to form a higher-definition image signal, the device comprising a first forming means that forms an interpolation signal from a signal of the same field, and an interpolation signal from a signal of a previous field. a second forming means for forming an interpolated signal from a rear field signal; a fourth forming means for forming an interpolated signal from a front field signal and a rear field signal; Calculate the difference signal between adjacent fields between the signal of the previous field preceding the field of interest to be formed, the signal of the subsequent field following the field of interest, and the signals spatially located above and below the interpolation scanning line. and determining means for determining the motion of an image, and appropriately selecting one of the interpolation signals based on the determination result.