KR930000953B1 - Circuit for reproducing tv-signals - Google Patents

Abstract

No content.

Description

TV signal reproduction circuit

1 is a block diagram of a first embodiment of a preferred apparatus for implementing the present invention.

2 is a view showing an image for explaining the principle of the present invention.

3 is a block diagram of a second embodiment of a preferred apparatus for implementing the present invention.

4 is a block diagram of a conventional apparatus.

5 is a view showing an image for explaining the operation of FIG.

* Explanation of symbols for main parts of the drawings

10: analog digital (A / D) converter 12, 40, 44: frame memory

14, 42, 46: motion compensation circuit 16: switch

18: sampling frequency converter 20: insertion circuit between fields

22: In-field insertion circuit 24: Sampling frequency converter

26: Differential amount low pass filter (LPF) 28: Absolute value circuit

30, 56: comparison circuit 32: dress hold value circuit

34: sampling frequency converter 36: switch

38: digital analog (D / A) converter 48: endgate

50: switch 52: differential amount erase circuit

54: minimum value circuit 58: linear mixed circuit

N: current frame N-1, N + 1: comparison frame

V: Area discrimination signal

The present invention relates to a television signal reproducing circuit which detects a motion signal of a video and inserts a television signal when the television video signal is reproduced.

One method of band compression of a television signal is between fields; And multiple subsample transmission schemes using offset subsampling between frames.

For example, the MUSE (Multiple Sub-Nyquist Sampling Encoding) method, which is the current high-definition television signal multiple sub-sample transmission method, can effectively realize band compression.

In this multi-subsample transmission scheme, the amount of motion of the current frame and the comparison frame preceding the current frame is detected in time, and the mixed ratio continuously changed in accordance with the amount of detection detects motion and moving pictures. Hwa) was mixed to obtain a reproduced image.

Next, with reference to Figs. 4 and 5, a reproduction circuit of the conventional multi-subsample transmission method will be described.

For simplicity, a still image is selected when the amount of motion is less than or equal to the appropriate threshold value, and motion is selected otherwise. A reproduced image is obtained by switching between these still and motion changes.

In FIG. 4, the analog-to-digital (A / D) converter 10 samples an analog television input signal at 16.2 MHz and converts it into a digital signal.

Blocks 12 to 20 are circuits of the stationary system.

The frame memory 12 stores a digital television signal from the A / D converter 10 and generates a frame delay signal.

The motion correction circuit 14 performs motion correction on the signal from the frame memory 12 in accordance with the motion vector.

The switch 16 alternately selects the output signals of the A / D converter 10 and the motion correction circuit 14 in alternating 32.4 MHz to perform interframe insertion.

The sampling frequency converter 18 converts the sample frequency of the signal from 32.4 HMz to 48.6 MHz from the switch 16 inserted between the frames.

The interfield insertion circuit 20 inserts the output signal of the sampling frequency converter 18 between the fields.

On the other hand, blocks 22 to 24 are circuits of the motion system.

The in-field insertion circuit 22 inserts the output signal of the A / S converter 10 into the field.

The sampling frequency converter 24, like the sampling frequency converter 18, converts the sampling frequency of the output signal of the in-field insertion circuit 22 from 32.4 MHz to 48.6 MHz.

Blocks 26 to 34 are circuits of the motion detection system.

The difference amount low pass filter (LPF) 26 receives a signal one frame ahead corrected according to the motion vector by the motion correction circuit 14 and a signal of the current frame from the A / D converter 10. And outputs a signal of one frame difference by band-limiting to 4 MHz which does not include a return component between frames.

The absolute value circuit 28 finds the absolute value of the output signal of the difference amount LPF 26.

The comparison circuit 30 compares the output signal A of the absolute value circuit 28 with the threshold value B from the threshold value 32.

For example, in the case of the absolute value signal A > the dress holder value B, the comparison circuit 30 outputs a motion signal of " 1 " as the movement region.

In other cases, the comparison circuit 30 outputs the motion signal "0" to the stop area.

The sampling frequency converter 34 converts the sampling frequency of the output signal of the comparison circuit 30 from 32.4 MHz to 48.6 MHz and outputs a motion signal.

The switch 36 selects the output signal of the still image system (inter-field insertion circuit 20) and the output signal of the motion system [sampling frequency converter 24] in accordance with the motion signal from the motion detection system, Switch and mix the signal with the motion system signal.

The mixed signal is returned to the analog signal by the digital analog (D / A) converter 38.

The motion area in the case of decoding a television signal according to the prior art of FIG. 4 will be described with reference to FIG.

In FIG. 5, (a) shows the comparison frame one frame before the current frame, that is, the image of the N-1th frame, and (b) shows the image of the current frame, that is, the Nth frame. Indicates.

In these comparison frames N-1 and the current frame N, areas A and C are still areas, and areas B and D are motion areas.

That is, in this example, the moving area is changed from the rhombic area B to the original area D during the time elapsed from the comparison frame N-1 to the current frame N.

In the effective range, the still areas A and C excluding the rhombus or the circle are, for example, a background and are parts which do not change.

In the prior art of FIG. 4, as described above, the result of the correlation calculation based on the absolute value of the difference component between the frames of the video signal corresponding to the video of the comparison frame N-1 and the current frame N is determined as a certain draw hole value. As a reference, the value is used as a motion signal.

The result by this correlation is shown in FIG.

In FIG. 5C, since the comparison frame N-1 is vector corrected by the motion correction circuit 14, the corrected regions corresponding to the regions A and B before correction are A 'and B '.

Significant correlation is detected in the region A 'and C where the still image area A' (vector corrected) of the comparison frame N-1 and the still image area C of the current frame N overlap. ) And the absolute value of the difference component between the frame of the video signal of the current frame N is equal to or less than the threshold value of the frame].

On the other hand, the region B '+ D occupied by at least one of the movement region B' (vector corrected) of the comparison frame N-1 and the movement region D of the current frame N is not detected and a significant correlation is not detected. (N-1), the difference between the current frame (N) and the video signal is larger than the threshold value of the threshold].

In FIG. 4 and FIG. 5, only motion detection by one-frame correlation is described. However, the same applies to two-frame correlation.

The prior art shown in FIGS. 4 and 5 is only one frame difference detection between two frames.

Therefore, the correlation calculation result based on the absolute difference value between the frames determines that the oblique area B 'and D in Fig. 5C are also the motion area of the current frame.

However, as shown in (b) of FIG. 5, this diagonal area B'D is not actually a moving area.

That is, in the conventional apparatus, although it is actually a still region, there is a drawback that it is judged as the movement region of the current frame.

In addition, the same shortcomings arise when converting from interlaced scanning to sequential scanning in the IDTV (Improved Definition TV) of the current standard system NTSC signal lamp.

It is therefore an object of the present invention to solve the above-mentioned drawbacks of the prior art and to provide a television signal reproduction circuit for inserting a television signal in response to a motion signal capable of more faithfully detecting a motion area.

The present invention relates to a television signal reproduction circuit for decoding a television signal, comprising a plurality of comparison frames before and after m frames (m: natural numbers), which are composed of an intra-field insertion circuit for obtaining a moving signal and an internal circuit for obtaining a still image signal. A plurality of motion candidate signals are detected by correlation calculation such as the absolute value of the difference component between the frames, and the motion signal 1 (V1) of the current frame is obtained from the logical product, the minimum value, the average value, etc. of the plurality of motion candidate signals. The motion signal 2 (V2) is obtained from one of the plurality of motion assistance signals or from a comparison value, an average value, and the like, and a plurality of comparison frames before and after m frames are mixed and inserted between the frames corresponding to the motion signal 2 (V2). Obtain the cellification signal. According to the obtained motion area, the still picture and the moving picture are mixed in correspondence with the motion signal 1 (V1).

In the present invention, unlike the prior art, the comparison frame is not only the frame immediately before the current frame but also the current frame and a plurality of frames before and after the m frames (m = 1, 2, 3, ...).

Thus, a plurality of motion candidate signals are obtained.

Since the motion area in the current frame is detected more faithfully by obtaining a motion area from these plurality of motion candidate signals, the area discrimination signal is more realistic.

Therefore, the internal insertion of the television signal is performed more faithfully.

EXAMPLE

1 to 3, a preferred embodiment of the television signal reproduction circuit of the present invention will be described.

In addition, the same block as FIG. 4 is shown with reference numeral, and abbreviate | omits description and a different block is demonstrated in detail.

In FIG. 1, the A / D converter 10 samples an analog television input signal and converts it into a digital signal.

In the still picture system, the combination of the frame memory 12 and the motion correction circuit 14 is the same as in the case of FIG.

However, in the present invention, the output signal of the A / D converter 10 is stored in the frame memory 40 and the signal delayed by one frame is supplied to the frame memory 12.

Therefore, when the output signal of the frame memory 40 is the current frame N, the output signal of the frame memory 12 becomes the comparison frame N-1 of one frame in time from the current frame N, and A / The output signal of the D converter 10 becomes the comparison frame N + 1 one frame after the current frame N in time.

The motion correction circuit 46 corrects the motion vector of the signal of the comparison frame N + 1 with respect to the current frame N.

As described above, since the output signal of the frame memory 40 corresponds to the current frame N, in the motion system, the in-field insertion circuit 22 and the sampling frequency converter 24 output the output signal of the frame memory 40. The same process as in the case of FIG.

Blocks 26 to 32 of the motion detection system are the same as in the case of FIG. 4 and operate in the same manner.

However, the output signal N of the comparison circuit 30 is stored in the frame memory 44, delayed by one frame, and supplied to the motion correction circuit 46.

The motion correction circuit 46 corrects the motion vector such that the position of the motion candidate signal M 'is referred to the comparison frame N as a reference.

Therefore, the output signal M 'of the motion correction circuit 46 calculates the result of correlation calculation based on the absolute value of the difference component between the frame of the current frame N and the comparison frame N-1. The motion candidate signal is converted into a value based on the reference value.

The output signal M of the comparison circuit 30 is a motion candidate signal obtained by correlation calculation based on the absolute value of the difference component between the current frame N and the frame of the comparison frame N + 1.

The AND gate 48 calculates the logical product of the motion candidate signal M and M '.

The sampling frequency converter 34 converts the sampling frequency of the output signal of the AND gate 48 from 32.4 MHz to 48.6 MHz to output the motion signal 1 V1 of the current frame M.

A process of obtaining the motion signal V1 of the current frame N will be described with reference to FIG.

In FIG. 2, (a) shows the video of the comparison frame one frame before the current frame, that is, the N-th frame, and (b) shows the image of the current frame, that is, the Nth frame. (C) shows a comparison frame one frame after the current frame, that is, the image of the N + 1th frame.

The areas A, C and E in these comparison frames N-1, (N + 1) and the current frame N are still areas, and areas B, D and F are motion areas.

That is, in this embodiment, in the time elapsed from the comparison frame (N-1) to the current frame (N) and the comparison frame (N + 1), the moving area is a rhombic area B in the circle area D. And it is changing to the area | region F of a triangle.

In addition, in the effective range, the stationary area excluding a rhombus, circle or triangle (A, C and E are for example, a background, which does not change).

In the present invention, first, a correlation calculation such as the absolute value of the difference component between the frame of the video signal corresponding to the video of the comparison frame N-1 and the current frame N is performed, and the motion candidate signal M 'is obtained.

The result by this correlation calculation is shown in FIG. 2 (d).

In addition, A 'and B' show that the area | region A and B were motion-corrected according to the motion vector.

In FIG. 2 (d), the region A'C (the logical product of the regions A 'and C) where the still image region A' of the comparison frame N-1 and the still image region C of the current frame N overlaps is meant. The correlation is detected and the motion candidate signal M 'is outputted as 0 (because the difference between the signals of the comparison frame N-1 and the current frame N is equal to or less than the threshold value).

On the other hand, an area B '+ D (logic of areas B' and D ') occupied by at least one of the movement area B' of the comparison frame N-1 and the movement area D of the current frame N means. No correlation is detected (the difference between the signals of the comparison frame N-1 and the current frame N is greater than the threshold value, etc., between the frames), and is determined as a motion region candidate. (M ') is output as 1.

Therefore, the content of FIG. 2 (d) becomes an output signal of the motion correction circuit 46. As shown in FIG.

Next, correlation calculation is performed such as the absolute value of the difference component between the frames of the video signal corresponding to the video of the comparison frame N + 1 and the current frame N and the like.

The result by this correlation calculation is shown in FIG. 2 (e).

In addition, E 'and F' show that the area | regions E and F were moved according to the motion vector.

In FIG. 2E, the region C · E (the logical product of the regions C and E ') where the still region E' of the comparison frame N + 1 and the still region C of the current frame N overlap is meant. The correlation is detected, and it is determined as the still area, and the motion candidate signal M is output as zero.

On the other hand, a significant correlation is detected in the movement region F 'of the comparison frame N + 1 in the region D + F' (the logic of the regions D and F ') occupied by at least one of the movement regions D of the current frame N. The motion candidate signal M is output as 1 without being determined as a motion area candidate.

Corresponds to the output signal of the content-sound comparison circuit 30 of FIG.

As shown in (d) and (e) in FIG. 2, in the correlation calculation such as the absolute value of the difference component between frames between the current frame and one comparison frame, the area B of the diagonal portion is the same as in the prior art. '· D' (OR product of region B 'and other than region D) and D · F' (OR product of region D and region F ') are motion region candidates, even though they are still regions in the current frame N As a result, the motion candidate signal M is output as one.

However, in the present invention, since the AND product of the motion candidate signals of (d) and (e) is obtained by the AND gate 48, the result is as shown in FIG.

That is, only the motion area D of the current frame N is detected as the motion area, and the motion signal 1 (V1) is output as one.

Therefore, it will be understood that the motion region and the motion signal 1 (V1) can be more faithfully detected. See FIG. 1 again.

In the still picture system, the switch 50 uses the frame memory 44 to delay the motion candidate signal M ', that is, a signal one frame before the current frame (current signal) N in accordance with the motion signal 2 (V2). Either one of (N-1) or the signal (N + 1) one frame later is selected.

For example, when the motion signal 2 (V2) is "1", the signal N + 1 is selected, and when the motion signal 2 (V2) is "0", the signal N-1 is selected. do.

The switch-16 alternately selects the selected signal and the current signal N from the current signal 40 and alternately 32.4 MHz from the frame memory frame memory 40 according to the switch 50, and inserts them between the frames.

The output signal of the switch 16 is processed in the same manner as in the case of FIG. 4 by the sampling frequency converter 18 and the insertion arc 20 between the fields.

The switch 36 selects the output signal of the still image system (inter-field insertion circuit 20) and the output signal of the motion system (sampling frequency converter 24) in accordance with the motion signal 1 (V1) from the motion detection system. Then, the stationary system signal and the motion system signal are switched and mixed.

The mixed signal is returned to the analog signal by the digital analog (D / A) converter 38.

In the above embodiment, the switch 50 is switched in accordance with the motion candidate signal M 'delayed by the frame memory 44, but is switched in accordance with the motion candidate signal M, which is an output signal of the comparison circuit 30. You may also do it.

That is, when the motion candidate signal M is the motion signal 2 V2 and the motion signal 2 V2 is "1", the signal N-1 is selected, and the motion area candidate M is "0." Is selected, the signal (N + 1) is selected.

Also, the motion signals 1, 2 (V1) and V2 of the current frame were obtained using the current frame and the comparison frame before and after one frame, but the current frame was compared from the comparison frame before and after the m frame (m = 1, 2, 3…). The interpolation of the signal can be performed faithfully even when the motion signals 1 and 2 (V1) and V2 of the frame are obtained.

Next, a case where linear mixing of still images and motions is performed in accordance with the motion amount will be described with reference to FIG.

In addition, the same block numbers as those in FIG. 1 are denoted by the same reference numerals, and descriptions of other blocks will be described in detail.

The output signal of the absolute value circuit 28 and the draw holder value from the draw holder value circuit 32 are supplied to the circuit 52 after removing the difference amount.

That is, when the output signal of the absolute value circuit 28 is larger than the threshold value, "0" is output when the value according to the subtraction result is small.

By eliminating this difference, the circuit 52 reduces the output of unnecessary difference components in the stationary region due to noise in the signal.

The motion candidate signal M and the motion candidate signal M 'through the frame memory 44 are supplied to the minimum value circuit 54, and the output signal thereof is supplied through the sampling frequency converter 34 to the motion signal 1 V1. Outputs

Further, the motion candidate signals M and M 'are supplied to the comparison circuit 56, and a motion signal 2 (V2) is obtained. The switch 50 is switched in accordance with the signal V2.

That is, the signal N-1 is selected when the motion candidate signal M is larger than (M '), and the signal N + 1 is selected when the motion candidate signal M is smaller than (M').

In addition, the linear mixing circuit 58 linearly mixes the output signal of the still system [inter-field insertion circuit 20] and the output signal of the motion system [sampling frequency converter 24] according to the motion signal 1 (V1). .

In this way, a signal obtained by linearly mixing still images and moving pictures is obtained in accordance with the motion signal 1 V1 of the current frame N. FIG.

In addition, in the case of linear mixing of inter-frame insertion, still picture, and moving picture according to the average value of a plurality of motion auxiliary signals, in Fig. 3, the minimum value circuit 54 and the comparison circuit 56 are composed of an average circuit. What is necessary is just to comprise the switch 50 with a linear mixing circuit.

According to the present invention as described above, a plurality of comparisons before and after the current frame m frame (m: natural number) are composed of an intra-field insertion circuit which receives a multi-sub sample transmission signal and obtains a moving picture signal and an insertion circuit between frames which obtain a still picture signal. A plurality of motion candidate signals are detected by a correlation calculation such as an absolute value of the difference component between the frames and the frames, and the motion signal 1 (V1) of the current frame is derived from the logical product, the minimum value, the average value, etc. of the plurality of motion candidate signals. The motion signal 2 (V2) can be obtained from either of the plurality of motion candidate signals or from a comparison value, an average value, and the like.

Therefore, the possibility of detecting a still image area as a moving picture area is lowered and the moving picture area can be detected more accurately, which is more accurate than the motion signal 1 (V1).

Further, according to the motion signal 2 (V2), a plurality of comparison frames before and after m frames (m: natural numbers) can be adapted and mixed to obtain the still picture signal for inserting frames.

Furthermore, the still image and the moving image are mixed according to the motion signal 1 (V1), so that the image quality can be improved as a whole.

In addition, the present invention can also be effectively used for IDTV according to the current NTSC method.

Claims (1)

A television signal reproduction circuit for decoding a television signal, comprising: means (40) for outputting a current frame; Means (12) and (42) for deriving a comparison frame signal before at least one frame and a comparison frame signal after at least one frame with respect to the current frame signal; Moving picture processing means (22), (24) for outputting moving pictures by inserting the current frame signal to the current frame signal output means (40) into a field; Still picture processing means (50), (16), (18), and (20) for adaptively mixing at least one frame before the frame derived by said comparison frame derivation means, inserting between frames, and outputting a still picture; In response to the frame signal in the comparison after the current frame signal and at least one frame, a first motion candidate signal M is generated according to a correlation by an absolute value of a frame difference difference component between the current frame signal and the frame signal after at least one frame. A second motion candidate signal M 'according to a correlation by an absolute value of a frame difference difference component between the current frame signal and the frame signal at least one frame before, in response to the current frame signal and the comparison frame signal at least one frame before. Means (26), (28), (32), (30), (44), (46), (52) for generating a); Means (48), (34), (54), (34) for deriving a motion signal (V1) in response to the first motion after signal and the second motion candidate signal; Means (56) for generating a second motion signal (V2) in response to one M 'of the first and second motion candidate signals or in response to the first and second motion candidate signals; Still image processing from the still image processing means in response to the first motion signal V1, without adaptive mixing of the comparison frame of the still image processing means in response to the second motion signal V2, and the moving picture processing means. And a television signal reproducing circuit having means for mixing moving pictures from the picture.

Images