JPH0851597A - Decompression device for band compression image signal - Google Patents

Abstract

PURPOSE:To save the memory capacity for a motion picture and still picture processing circuit by selecting plural consecutive video signals obtained by applying inter-frame interpolation processing to a high definition TV signal for each field and using the video signal and applying inter-field processing to the signal. CONSTITUTION:A digital signal obtained by converting a received MUSE signal by an A/D converter 11 is processed sequentially by a selection circuit 109 and field (Fi) memories 110, 111 in an inter-frame interpolation (FI) circuit 13. Furthermore, the circuit 13 uses an inter-frame moving vector correction circuit 120 to provide an output of FI signals N0 to N2 to an inter-Fi processing circuit 20 for each Fi by selection circuits 114, 115 alternately. Each of luminance chrominance signal processing circuit 1100 to 1400 in the processing circuits 24, 25 in the moving and still area of the circuit 20 process the input signal according to a control signal and outputs each signal required to decode an output RGB signal. Thus, the memory required for the circuits 24, 25 is used in common by the memories 110, 111 to save the memory capacity.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a band-compressed image signal decompression device for decompressing an image signal band-compressed by multiple sub-Nyquist sampling, and more particularly to a memory required for inter-field processing in frame units. The present invention relates to a decompression device for a band-compressed image signal capable of reducing the capacity.

Further, depending on the amount of movement of the input image signal, the inter-field processed moving image signal and the intra-field processed moving image signal are mixed at a predetermined ratio to improve the image quality of the moving image. It relates to the device.

[0003]

2. Description of the Related Art MUSE (Mult) using multiple sub-Nyquist sampling is one of the methods of compressing and transmitting a wide band high-definition television signal into a practical level band that can be transmitted.
iple Sub-Nyquist Sampling Encoding) method (Ninomiya et al .: "High-definition TV satellite 1-channel transmission method", Technical Report of the Television Society of Japan, TEBS95-2, Vol.7, No.44),
Test broadcasting using satellite broadcasting has already been conducted. MU
In the SE method, the movement of a subject in an image is detected by using a motion detection circuit, and a region in which the subject is moving (hereinafter referred to as a moving region) and a region in which the subject is stationary (hereinafter referred to as a static region) are described. ) And by performing different processing respectively, the bandwidth is about 2
A wide band video signal of 0 MHz is compressed to about 8 MHz. Moreover, in recent years, a method for improving the image quality of the MUSE method (Koshi et al .: "Study on video processing in the MUSE method", IEICE Journal, Vol.75J-BI, No.10, pp639-646, (1992),
Or Koshi et al .: "Study on color difference signal processing of MUSE method", IEEJ, Vol.76J-BI, No.3, pp290-298, (199
3), or Koshi et al .: "High image quality of MUSE", IEICE, 137th workshop, pp17-20 (1993)). (Hereinafter, this proposed method will be referred to as an image quality improvement type MUSE method. On the other hand, the conventional method will be referred to as a current MUSE method.) This image quality improvement type MUSE method uses a frame unit for moving region and still region processing. Image quality is improved by introducing inter-field processing. The operation of the image quality improving MUSE method will be described below.

An image quality improvement type MUSE type band compression device is composed of, for example, a circuit shown in FIG. In FIG. 27, the wideband RGB signal supplied to the input terminal 1 is
The AD converter 11 converts the digital signal at a sampling frequency of 48 MHz and supplies the digital signal to the matrix circuit 102. In the matrix circuit 102, the RGB signals are converted into a luminance signal (Y) and color difference signals (BY, RY), respectively. The converted luminance signal is supplied to the moving area luminance signal processing circuit 200 and the still area luminance signal processing circuit 300, and the color difference signal is supplied to the color signal processing circuit 400.

The moving area luminance signal processing circuit 200 is composed of a circuit as shown in FIG. 28, for example. The spectral structures of signals at points a to f in the figure are shown in FIGS. The signals A and B correspond to the signals at points A and B in FIG. Input brightness signal is about 20MH
Although it has a z band, it is band-limited by the 16 MHz horizontal low-pass filter 201 and becomes an output signal having a spectral structure as shown in FIG. The output signal of the 16 MHz horizontal low-pass filter 201 is supplied to the sampling frequency conversion circuit 202, and the sampling frequency changes from 48.6 MHz to 3 MHz.
Converted to 2.4 MHz. The signal with the converted sampling frequency is supplied to the two-dimensional low-pass filter 203. Two
In the dimensional low-pass filter 203, the band in the diagonal direction is limited, and the output signal has a spectral structure as shown in FIG. The output signal of the two-dimensional low-pass filter 203 is 8
It is supplied to the MHz horizontal low-pass filter 204 and the subtractor 205.
The 8 MHz horizontal low-pass filter 204 extracts horizontal frequency components of 8 MHz or less, and becomes an output signal having a spectral structure as shown in FIG. 29 (c). 2 in subtractor 205
-Dimensional low-pass filter 203 output signal and 8 MHz horizontal low-pass filter 204 output signal difference, horizontal 8 MHz or higher high-frequency component is obtained, the output signal having a spectral structure as shown in FIG. Becomes The output signal of the subtractor 205 is supplied to the field memory 207, and the output signal of the field memory 207 is supplied to the field memory 208. The output signal of the subtractor 205 and the output signal of the field memory 208 are supplied to the selection circuit 209, and either one of the fields is selected and output alternately according to the control signal from the control signal input terminal 100. The output signal of the field memory 207 and the output signal of the selection circuit 209 are supplied to the inter-field vertical low pass filter 210. Here, the output signal of the selection circuit 209 is 1 with respect to the output signal of the field memory 207.
There is a relationship before or after the field period. That is, inter-field processing is performed in frame units (hereinafter, this is referred to as frame completion). In the inter-field vertical low-pass filter 210, the inter-field vertical low-pass filter processing is performed on the high-frequency signal of 8 MHz or more in the horizontal direction.
The output signal has a spectral structure as shown in (e). In the adder 211, the addition processing of the output signal of the 8 MHz horizontal low-pass filter 204 and the output signal of the inter-field vertical low-pass filter 210 delayed by one field period for timing adjustment by the field memory 206 is performed, and FIG.
The signal has a spectrum structure as shown in (f).
The output signal of the adder 211 is supplied to the mixing circuit 103 as a moving area luminance signal processing output.

The static area luminance signal processing circuit 300 is composed of, for example, the circuit shown in FIG. In the figure, signals C and D correspond to the signals at points C and D in FIG. 27, the input luminance signal is supplied to the field memory 301, and the output signal of the field memory 301 is supplied to the field memory 302. The input luminance signal and the output signal of the field memory 302 are supplied to the selection circuit 303, and either one of the fields is selected and output alternately by the control signal from the control signal input terminal 100 for each field. The output signal of the field memory 301 and the output signal of the selection circuit 303 are supplied to the inter-field two-dimensional low pass filter 304. Here, the output signal of the selection circuit 303 is 1 with respect to the output signal of the field memory 301.
There is a relationship before or after the field period. The inter-field two-dimensional low-pass filter 304 performs pre-filtering by inter-field processing by frame completion.
The output signal of the inter-field two-dimensional low-pass filter 304 is decimated in sampling frequency from 48 MHz to 24 MHz in the field offset sub-sampling circuit 31, and 1
It is supplied to the 2 MHz horizontal low-pass filter 305. 12 MH
The z horizontal low-pass filter 305 performs interpolation processing to obtain a signal having a sampling frequency of 48 MHz again. 12 MHz
The output signal of the horizontal low-pass filter 305 is supplied to the sampling frequency conversion circuit 306, and the sampling frequency is 4
Converted from 8.6 MHz to 32.4 MHz. The signal whose sampling frequency has been converted to 32.4 MHz is supplied to the mixing circuit 103 as a still area luminance signal processing output.

The color signal processing circuit 400 is composed of a circuit as shown in FIG. 31, for example. The spectral structures of signals at points a and b in the figure are shown in FIGS. 32 (a) and 32 (b). The terminals E, F, G correspond to the portions E, F, G in FIG. Input color difference signals (BY, RY) are 8 MHz at 8 MHz horizontal low-pass filters 401 and 404, respectively.
The band is limited below, and the output signal has a spectral structure as shown in FIG. The output signals of the 8 MHz horizontal low-pass filters 401 and 404 are supplied to the sampling frequency conversion circuits 402 and 405, respectively, and the sampling frequency is converted from 48.6 MHz to 16.2 MHz. The signal with the converted sampling frequency is filtered by the vertical low-pass filter.
The signals are supplied to 403 and 406 respectively, the band is limited in the vertical direction, and line-sequential multiplexing is performed in the line-sequential multiplexing circuit 407.
The spectrum of the line-sequential multiplexed signal is as shown in FIG. The line-sequential multiplexed signal is processed by the moving area color signal processing circuit 50
0 and the static area color signal processing circuit 600.

The moving area color signal processing circuit 500 is shown in FIG.
It is composed of a circuit as shown in FIG. The spectral structure of the signal at each point a and b in the figure is shown in FIGS. The terminals G and H correspond to the portions G and H in FIG. The output signal from the line-sequential multiplexing circuit 407 is supplied to the field memory 502, and the output signal of the field memory 502 is supplied to the field memory 503. Selection circuit 5
The output signal of the line-sequential multiplexing circuit 407 and the output signal of the field memory 503 are supplied to 04, and the control signal input terminal 1
Either one of the fields is selected and output alternately by a control signal from 00. The output signal of the field memory 502 and the output signal of the selection circuit 504 are supplied to the inter-field vertical low pass filter 505. Where the field memory 502
The output signal of the selection circuit 504 has a relationship before or after one field period with respect to the output signal of. The inter-field vertical low-pass filter 505 performs band limitation in the vertical direction by inter-field processing due to frame completion.
The signal has a spectral structure as shown in (b).
The sampling frequency is 16.2 in the field offset sub-sampling circuit 51 for the band-limited signal.
Converted from MHz to 8.1 MHz. The field offset sub-sampled signal is supplied to the mixing circuit 104 as a moving area color signal processing output.

The static area color signal processing circuit 600 is composed of, for example, the circuit shown in FIG. In the figure, terminals G and I are shown in FIG.
It corresponds to the G and I parts of 7. Line-sequential multiplexing circuit 40
The line-sequential signal from 7 is supplied to the field offset sub-sampling circuit 601, and the sampling frequency is 1
It is sub-sampled from 6.2 MHz to 8.1 MHz. The field offset sub-sampled signal is supplied to the mixing circuit 104 as a still region color signal processing output.

In the mixing circuit 103, the moving area luminance signal circuit 200
And the output signal of the stationary area luminance signal processing circuit 300 are mixed at a predetermined ratio according to the motion detection signal from the motion detection circuit (not shown) supplied to the control signal input terminal 101. Similarly, in the mixing circuit 104, the output signal of the moving area color signal processing circuit 500 and the still area color signal processing circuit 6
The output signal of 00 is mixed at a predetermined ratio according to the motion detection signal. In the TCI circuit 105, the color difference signal from the mixing circuit 104 is time-compressed to 1/4 and is converted to a sampling frequency of 32.4 MHz, and at the same time, the mixing circuit 10
The luminance signal from 3 is multiplexed in the blanking period and T
Obtain the CI signal. The sampling frequency of the TCI signal is 32.
Converted from 4MHz to 16.2MHz, DA converter 2
It is converted into an analog signal at 1 and analog M from the output terminal 2
It is output as a USE signal.

Next, the MUS band-compressed to about 8 MHz by the above-described image quality improving MUSE type band compression device.
A restoration device that restores the E signal to the original wideband signal will be described.

FIG. 36 shows the construction of an image quality improvement type MUSE restoration apparatus. In FIG. 36, the input terminal
The analog MUSE signal supplied to 1 is the AD converter 11
Is converted into a digital MUSE signal having a data rate of 16.2 MHz and then supplied to the interframe interpolation circuit 13. The inter-frame interpolation circuit 13 performs inter-frame interpolation according to the sub-sample phase information supplied to the control signal input terminal 106 to obtain an inter-frame interpolation signal of 32.4 MHz rate.

The inter-frame interpolation signal is supplied to a moving area luminance signal processing circuit 700, a moving area color signal processing circuit 800, a still area luminance signal processing circuit 900, and a still area color signal processing circuit 1000, respectively.

The moving area luminance signal processing circuit 700 is composed of, for example, the circuit shown in FIG. 38 (a) to (e) show the spectral structures of signals at points a to e in the figure.
The signals A and B correspond to the signals at points A and B in FIG. In FIG. 37, the input inter-frame interpolation signal is subjected to field interpolation processing in the two-dimensional low pass filter 701. In this case, only the signal of the current field is extracted from the inter-frame interpolation signal, and the field interpolation processing is performed. The field interpolation output has a spectral structure as shown in FIG. Two-dimensional low pass filter
Output signal of 701 is 8MHz horizontal low-pass filter 702 and subtractor
It is supplied to 703. In the 8MHz horizontal low pass filter 702,
A frequency component of 8 MHz or less in the horizontal direction is extracted, and FIG.
The output signal has a spectral structure as shown in (b). The subtracter 703 obtains the difference between the output signal of the two-dimensional low-pass filter 701 and the output signal of the 8 MHz horizontal low-pass filter 702 to obtain a high-pass signal of 8 MHz or more in the horizontal direction, and the spectrum structure as shown in FIG. 38 (c). Is an output signal having. The output signal of the subtractor 703 is supplied to the field memory 705,
The output signal of the field memory 705 is the field memory 706.
Is supplied to. In the selection circuit 707, the output signal of the subtractor 703 and the output signal of the field memory 706 are input, and either one of the fields is selected and output alternately by the control signal from the control signal input terminal 107 for each field. However, they are selected so as to have the same phase relationship as the inter-field processing on the side of the band compression device. The output signal of the field memory 705 and the output signal of the selection circuit 707 are vertical low-pass filter 7 between fields.
Supplied on 08. Here, the output signal of the selection circuit 707 has a relationship before or after one field period with respect to the output signal of the field memory 705. The inter-field vertical low-pass filter 708 subjects the horizontal high-frequency signal of 8 MHz or more to vertical low-pass filtering between fields, and becomes an output signal having a spectrum structure as shown in FIG. 38 (d). In the adder 709, the output signal of the 8 MHz horizontal low-pass filter 702 delayed by one field for the timing adjustment by the field memory 704 and the output signal of the inter-field vertical low-pass filter 708 are added, and FIG. The output signal has a spectral structure as shown in FIG. Adder 709
Is supplied to the sampling frequency conversion circuit 710, and the sampling frequency is changed from 32.4 MHz to 48.
Converted to 6MHz. The signal with the converted sampling frequency is supplied to the mixing circuit 112 in FIG.

The moving region color signal processing circuit 800 is shown in FIG.
The circuit shown in FIG. The spectral structures of the signals at points a and b in the figure are shown in FIGS. 40 (a) and 40 (b). The signals C and D correspond to the signals at points C and D in FIG. In FIG. 39, the input signal of the time expansion circuit 801 is a 32.4 MHz rate interframe interpolation signal in the TCI state. Therefore, the time expansion circuit 801 expands only the color signal that has been compressed 1/4 time to 4 times, and obtains the color signal interpolated between the frames with the sampling frequency of 8.1 MHz. Time extension circuit 801
Is output to the vertical low pass filter 802. The vertical low-pass filter 802 extracts only the signal of the current field from the inter-frame interpolation signal supplied, performs field interpolation by the vertical low-pass filter, and outputs the spectrum structure as shown in FIG. Get the signal. The output signal of the vertical low-pass filter 802 is supplied to the field memory 803, and the output signal of the field memory 803 is supplied to the field memory 804. Vertical low pass filter in select circuit 805
The output signal of 802 and the output signal of the field memory 804 are supplied, and either one of the fields is selected and output alternately by the control signal from the control signal terminal 107 for each field. However, it is selected so as to have the same phase relationship as the frame completion processing on the side of the band compression device. The output signal of the field memory 803 and the output signal of the selection circuit 805 are supplied to the inter-field vertical low pass filter 806. Here, the output signal of the selection circuit 805 has a relationship before or after one field period with respect to the output signal of the field memory 803. The inter-field vertical low-pass filter 806 performs inter-field interpolation processing by frame completion, reproduces high-frequency components of 4 MHz or more folded between fields, and outputs inter-field interpolated signals with a sampling frequency of 16.2 MHz. obtain. The interfield interpolation signal has a spectral structure as shown in FIG. The inter-field interpolation signal is supplied to the mixing circuit 113 in FIG.

The static area luminance signal processing circuit 900 is composed of, for example, the circuit shown in FIG. Signals E and F in the figure are shown in FIG.
6 corresponds to signals at points E and F. In FIG. 41, the input inter-frame interpolation signal is band-limited by the 12 MHz horizontal low-pass filter 901 and then supplied to the sampling frequency conversion circuit 902 so that the sampling frequency is 3
Converted from 2.4 MHz to 48.6 MHz. The output signal of the sampling frequency conversion circuit 902 is subsampled to a sampling frequency of 24.3 MHz by the inter-field offset sub-sampling circuit 907 and supplied to the field memory 903. The output signal of the field memory 903 is supplied to the field memory 904. In the selection circuit 905, the output signal of the sampling frequency conversion circuit 902 and the output signal of the field memory 904 are input, and either one of them is selected and output alternately for each field by the control signal from the control signal input terminal 107. However, it is selected so as to have the same phase relationship as the pre-filtering by frame completion on the compression device side. The output signal of the field memory 903 and the output signal of the selection circuit 905 are supplied to the inter-field two-dimensional low pass filter 906. Where the field memory 9
The output signal of the selection circuit 905 has a relationship before or after one field period with respect to the output signal of 03. Between fields 2
In the dimensional low-pass filter 906, inter-field interpolation by frame completion is performed by a two-dimensional low-pass filter between fields, and a still area signal is reproduced. The inter-field interpolation signal is supplied to the mixing circuit 112 in FIG.

The static area color signal processing circuit 1000 is composed of, for example, the circuit shown in FIG. Signals G and H in the figure are shown in FIG.
Corresponding to signals at points G and H. Similarly to the one described in the moving area color signal processing circuit, the time extension circuit 1001 of FIG. 42 performs the time extension of 4 times on the image signal which has been subjected to the time compression of 1/4, and the sampling frequency becomes 8 times. 1. Obtain inter-frame interpolated signal of 1 MHz. The output signal of the time extension circuit 1001 is supplied to the field memory 1002,
The output signal of the field memory 1002 is the field memory 10
Supplied to 03. In the selection circuit 1004, the output signal of the time extension circuit 1001 and the output signal of the field memory 1003 are input, and either one of them is selected and output alternately according to the control signal from the control signal input terminal 107 for each field. However, they are selected so as to have the same phase relationship as the inter-field processing on the side of the band compression device. The output signal of the field memory 1002 and the output signal of the selection circuit 1004 are supplied to the inter-field vertical low pass filter 1005. Here, the output signal of the selection circuit 1004 is 1 with respect to the output signal of the field memory 1002.
There is a relationship before or after the field period. The inter-field vertical low-pass filter 1005 performs inter-field interpolation processing based on frame completion to obtain a 16.2 MHz inter-field interpolation signal. Inter-field interpolation signal is shown in Fig. 3.
6 to the mixing circuit 113.

In the mixing circuit 112, the moving area luminance signal and the still area luminance signal are mixed at a predetermined ratio according to the motion detection signal supplied from the motion detection circuit (not shown) to the control signal input terminal 108. The mixed luminance signal is supplied to one input terminal of the inverse matrix circuit 5.

In the mixing circuit 113, the moving area color signal and the still area color signal are mixed at a predetermined ratio according to the motion detection signal from the control signal input terminal 108. The mixed color signal is supplied to the line-sequential decoding circuit 6. The color signals supplied to the line-sequential decoding circuit 6 are subjected to vertical interpolation processing to obtain (BY) and (RY) signals. These color difference signals have sampling frequencies from 16.2 MHz to 48.
After being converted to 6 MHz, it is supplied to the remaining input terminals of the inverse matrix circuit 5. In the inverse matrix circuit 5, an inverse matrix operation is performed from the luminance signal and the color difference signal,
It is restored to the original broadband RGB signal. These signals are converted into analog signals by the DA converter 21 and output.

[0020]

As described above, in the image quality improvement type MUSE method, for example, in the moving area luminance signal processing, as shown in FIG. 29 (f) or FIG. 38 (e), 8 MHz or more and 1125 or more. Attenuation of diagonal components of more than 2 TV lines to reduce aliasing interference,
As a result, the dynamic region luminance signal band is expanded. Furthermore, the frame completion processing is introduced to prevent 15 Hz flicker interference. As for the moving area color signal processing, by introducing inter-field processing by frame completion, the band of the moving area color signal is expanded to about 8 MHz as shown in FIG. , Realizes almost the same bandwidth. Furthermore, with regard to still area processing, the introduction of frame completion processing completely
Since it is possible to perform processing using pixels for fields and pixels of unnecessary fields are not mixed, the accuracy of motion detection is improved, and it is possible to prevent erroneous processing of a still area as a moving area.

As described above, according to the image quality improving MUSE system, it is possible to significantly improve the image quality of high-definition television broadcasting.

However, the image quality improvement type MUSE type restoration device requires a large number of field memories for obtaining field delay signals in order to perform inter-field processing due to frame completion, and is compared with the current MUSE type restoration device. There was a problem that it would lead to cost increase.

Further, the moving area luminance signal processing circuit of the current MUSE band compression apparatus uses only intra-field processing and does not introduce inter-field processing. As described above, the inter-field processing by frame completion, which is introduced in the image quality improving MUSE method, needs to be introduced on both the compression side and the decompression side. Therefore, the MUS by the current MUSE band compression device
When the E signal is restored by the image quality improvement type MUSE restoration device, there is a problem that interference such as ringing and folding occurs in the moving area.

Similarly, the interference caused when the MUSE signal compressed by the current MUSE band compressing device is restored by the image quality improving MUSE restoring device is more remarkable in the moving region color signal. As described above, in the moving region color signal processing of the image quality improvement type MUSE compression apparatus, the time resolution is limited to 15 Hz by the inter-field processing by frame completion. Therefore, no problem will occur even if inter-field processing by frame completion is performed on this signal in the image quality improvement type MUSE restoration apparatus. However, in the moving region color signal processing of the current MUSE band compressing apparatus, since the processing is all in the field, the time resolution is 30 Hz. Therefore, when such a signal is subjected to inter-field processing (interpolation) due to frame completion, there is a problem that multi-line interference occurs in a restored image. This becomes more pronounced in the fast motion of high saturation images.

The present invention solves the above-mentioned conventional problems, and a band-compressed image signal capable of reducing the memory capacity required for a moving area processing circuit and a still area processing circuit in which inter-field processing by frame completion is introduced. It is an object of the present invention to provide a restoration device of

Further, band compression capable of reducing ringing and aliasing interference of the moving area luminance signal which occurs when the MUSE signal compressed by the current MUSE type band compression device is restored by the image quality improvement type MUSE type restoration device. An object is to provide an image signal restoration device.

Further, multi-line interference generated at the edge portion of the moving region color signal, which occurs when the MUSE signal compressed by the current MUSE band compressing device is restored by the image quality improving MUSE restoring device, is reduced. An object of the present invention is to provide a decompression device for a band-compressed image signal.

[0028]

In order to achieve the above object, the first invention is a delay means for transmitting a video signal interpolated between frames by using a band-compressed high-definition television signal as an input. A signal selecting means is provided for selecting N0 and N2 for each field from the three time-sequential video signals N0, N1 and N2 obtained from the delay means, and the selecting means outputs the signal. The inter-field processing is configured to be performed using the video signal to be processed and the video signal of N1.

A second aspect of the invention is to input interframe interpolating means for performing interframe interpolating processing and output signals from the interframe interpolating means, and for three consecutive fields of N0, N1 and N2. Field delay means for extracting inter-frame interpolated signals, first selecting means for selectively outputting one of the N0 and N2, and second selecting means for selectively outputting one of the N0 and N2. It is configured to have and.

A third aspect of the present invention provides an inter-frame interpolation means for performing inter-frame interpolation processing while performing motion correction on an input band compressed signal based on motion vector information, and an output signal from the inter-frame interpolation means. Enter, N0, N1 and N
A field delay means for extracting two consecutive inter-frame interpolated signals for three fields, a selection means for selectively outputting one of N0 and N2, and an output signal from the selection means for the motion vector information. And a motion vector correcting means for performing motion correction based on the above.

The motion area processing means of the fourth aspect of the present invention is the first field interpolation means, the second field interpolation means, and the field interior output from the first field interpolation means. Vertical low-frequency component extraction means for extracting a vertical low-frequency component by inter-field operation using the interpolated signal and the field interpolated signal output from the second field interpolating means; Vertical for extracting vertical high frequency components by inter-field operation using the field interpolation signal output from the field interpolation means and the field interpolation signal output from the second field interpolation means High frequency component extraction means and horizontal low frequency component extraction means for extracting horizontal low frequency components from the vertical high frequency components extracted from the vertical high frequency component extraction means. And a adding means for adding the vertical low frequency component extracted by the vertical low frequency component extracting means and the horizontal low frequency component extracting by the horizontal low frequency component extracting means. There is.

A fifth aspect of the invention has a structure having a second moving area processing means for processing the moving area only by the in-field processing.

A sixth aspect of the present invention has a structure having a third moving area processing means for switching between intra-field processing and inter-field processing in accordance with the movement of an image.

According to a seventh aspect of the invention, at the output end of the moving area color signal processing means, a plurality of stages of median value extracting means for extracting the median value level from the pixels in the predetermined area are provided in series.

According to an eighth aspect of the invention, at the output end of the moving area color signal processing means, maximum value extracting means for extracting the maximum value from the pixels in the predetermined area and minimum value for extracting the minimum value from the pixels in the predetermined area. It has a structure including an extracting means.

[0036]

According to the first and second aspects of the present invention, the memory required for inter-field processing by frame completion can be shared by the memory used for inter-frame interpolation means, and the memory can be reduced.

According to the third invention, the motion vector correction circuit can be further shared, and the circuit can be significantly reduced.

According to the fourth invention, the circuit scale of the moving area luminance signal processing circuit can be reduced.

According to the fifth and sixth inventions, the current MU
Band-compressed image signal by SE method for improving image quality MUSE
It is possible to suppress ringing, aliasing interference, and the like that occur in a fast moving image or the like when the decoding is performed by the decoding device of the system.

According to the seventh and eighth inventions, the current MU
Band-compressed image signal by SE method for improving image quality MUSE
It is possible to remove multi-line interference that occurs in a fast-saturated high-saturation image when it is decoded by a decoding device of the system.

[0041]

【Example】

(Embodiment 1) An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the arrangement of a band-compressed image signal decompression device according to the first embodiment of the present invention.

In FIG. 1, the analog MUSE signal supplied to the input terminal 1 is converted into a digital MUSE signal having a data rate of 16.2 MHz by the AD converter 11, and then supplied to the interframe interpolation circuit 13. It The interframe interpolating circuit 13 includes a selection circuit 109 to which the digital MUSE signal is supplied to one input terminal, and a storage capacity to which the output signal of the selection circuit 109 is supplied and which can write the digital MUSE signal for two fields. And a field memory 111 having a storage capacity capable of writing the digital MUSE signal read from the field memory 110 for two fields.
And the MUS read from this field memory 111
The inter-frame motion vector correction circuit 120 performs motion correction for one frame on the E signal, and the signal from the inter-frame motion vector correction circuit 120 is selected by the selection circuit 109.
Is supplied to the other input terminal of. And the selection circuit
109 is controlled to be alternately switched based on the sub-sampling phase information supplied to the control signal input terminal 106, and the data rate from the output terminal is 32.4 MHz.
To obtain an enhanced MUSE signal.

That is, the inter-frame interpolation circuit 13 is
The MUSE data for a total of four fields is sequentially interpolated by the two field memories 110 and 111, and the selection circuit 109 selects the MUSE signal (current field signal) output from the AD converter 11. From the output terminal, a signal (signal N0) obtained by interpolating the signal of the current field and the signal of two fields before the frame is obtained, and the signal of one field and the signal of three fields before are framed from the field memory 110. Interpolated signal (Signal N1)
From the field memory 111, a signal (signal N2) obtained by interpolating a signal two fields before and a signal four fields before is obtained. Therefore, in the selection circuit 109, of the inter-frame interpolated signals of the two-field-preceding signal and the four-field-preceding signal output from the inter-frame motion vector correcting circuit 120, the four-field-preceding signal. And the signal of the current field is exchanged. Also,
The inter-frame motion vector correction circuit 120 performs motion correction on the input signal based on the motion vector signal supplied to the control signal input terminal 14. Through the above operation, the inter-frame interpolation circuit 13 can obtain the inter-frame interpolation signal subjected to the motion vector correction.

The interframe interpolation signals N0 and N2 obtained by the interframe interpolation circuit 13 are supplied to the selection circuits 114 and 115, respectively, and the output signals from the selection circuits 114 and 115 and the interframe interpolation signal N1 are supplied. Is supplied to the inter-field processing circuit 20. The inter-field processing circuit 20 comprises a moving area processing circuit 24 and a still area processing circuit 25, and the output signal of the selection circuit 114 and the inter-frame interpolation signal N1 are supplied to the moving area processing circuit 24. In addition, the selection circuit 115
The output signal and the inter-frame interpolation signal N1 are supplied to the stationary area processing circuit 25. Here, the selection circuits 114 and 1
Reference numeral 15 operates so as to switch for each field in accordance with the control signal supplied to the control signal input terminal 107.

The moving area processing circuit 24 uses the output signal of the selection circuit 114 and the interframe interpolation signal N1 to perform interfield processing in the moving area. The moving area processing circuit includes a moving area luminance signal processing circuit 1100 and a moving area color signal processing circuit 1200.

On the other hand, the stationary area processing circuit 25 includes a selection circuit 115.
The inter-field processing in the stationary area is performed by using the output signal of 1 and the inter-frame interpolation signal N1. The static area processing circuit is composed of a static area luminance signal processing circuit 1300 and a static area color signal processing circuit 1400.

The moving area luminance signal processing circuit 1100 is composed of, for example, the circuit shown in FIG. The spectral structures of the signals at points a to e in the figure are shown in FIGS. The signals A, B and C correspond to the signals at points A, B and C in FIG.

In FIG. 2, signal B is an interframe interpolated signal N of the signal one field before and the signal three fields before.
It is 1. Therefore, the two-dimensional LPF to which it is supplied
In 1104, only the signal one field before is extracted from the signal B, and field interpolation processing is performed using only that signal. On the other hand, as the signal A, the inter-frame interpolation signal N0 between the current field signal and the signal two fields before, or the inter-frame interpolation signal N2 between the signal two fields before and four fields before is alternately supplied for each field. ing. In the two-dimensional LPF 1101, when the former is supplied, the signal of the current field is extracted, and when the latter is supplied, only the signal of two fields before is extracted and subjected to field interpolation processing.
As a result, both output signals have the spectral structure shown in FIG. Here, the output signal of the two-dimensional LPF 1104 is the field interpolation signal of the 1-field previous signal, and the output signal of the two-dimensional LPF 1101 is the field interpolation signal of the current field / 2-field previous signal. Therefore, it is understood that inter-field processing by frame completion can be realized by performing processing between both.

The output signal of the two-dimensional LPF 1101 is subjected to inter-field motion vector correction in the inter-field motion vector correction circuit 1110, and then supplied to the 8 MHz horizontal low pass filter 1102 and the subtractor 1103. Here, the signal supplied to the inter-field motion vector correction circuit 1110 is the field interpolation signal of the current field signal or the field interpolation signal of the two-field previous signal. The correction is performed so as to match the image phase of the field interpolation signal of the 1-field preceding signal which is the output signal. That is, the correction direction is switched in association with the selection operation for each field in the selection circuit 114.

The inter-field motion vector correction circuit 1110 is composed of, for example, the circuit shown in FIG. In FIG. 4, the input signal is the output signal of the two-dimensional LPF 1101 and is supplied to the multistage line delay device 1150. The multi-stage line delay device 1150 delays the input signal on a line-by-line basis and supplies a plurality of different line delay signals to the switching circuit 1151. The switching circuit 1151 performs the switching operation according to the control signal from the selection circuit 1154. The output signal of the switching circuit 1151 is supplied to the shift register 11152. The shift register 1152 delays the input signal at a rate of 32.4 MHz and supplies the output signal to the switching circuit 1153 from a plurality of output terminals Q0 to Qn for each pixel. The switching circuit 1153 performs a switching operation according to the control signal output from the selection circuit 1155. The output signal from the switching circuit 1153 becomes the output signal of the inter-field motion vector correction circuit 1110. Here, the selection circuit 1154 has a control input terminal
The inter-field vertical motion vector signal supplied from 15a and the inter-field vertical motion vector signal whose sign is inverted via the sign inversion circuit 1156 are supplied. Therefore, by selectively outputting these signals for each field according to the control signal from the control signal input terminal 107, the vertical correction direction can be switched in conjunction with the selection operation for each field in the selection circuit 114 of FIG. .
Similarly, the inter-field horizontal motion vector signal supplied from the control input terminal 15b and the inter-field horizontal motion vector signal whose sign has been inverted via the sign inversion circuit 1157 are supplied to the selection circuit 1155. Therefore, by selectively outputting these signals for each field in accordance with the control signal from the control signal input terminal 107,
The horizontal correction direction can be switched in conjunction with the selection operation for each field in the selection circuit 114. As described above, according to the inter-field motion vector correction circuit shown in FIG. 4, it is possible to switch the correction direction in both vertical and horizontal directions.

On the other hand, the output signal of the two-dimensional low pass filter 1104 is supplied to the 8 MHz horizontal low pass filter 1105 and the subtractor 1106. 8 MHz horizontal low-pass filters 1102 and 1105 extract frequency components below 8 MHz horizontally, and both are shown in FIG.
The output signal has a spectral structure as shown in (b). In the subtractors 1103 and 1106, the difference processing between the output signal of the two-dimensional low-pass filter and the output signal of the 8 MHz horizontal low-pass filter is performed, and a high-pass signal of 8 MHz or more in the horizontal direction is obtained, both of which are shown in FIG. The output signal has such a spectral structure. The output signals of the subtractors 1103 and 1106 are supplied to the inter-field vertical low pass filter 1107. The inter-field vertical low-pass filter 1107 performs a vertical low-pass filtering process on a high-frequency signal of 8 MHz or more horizontally between fields, and then, as shown in FIG.
The output signal has a spectral structure as shown in. The adder 1108 performs addition processing on the output signal of the 8 MHz horizontal low-pass filter 1105 and the output signal of the inter-field vertical low-pass filter 1107 to obtain an output signal having a spectrum structure as shown in FIG. 3 (e). . That is, according to the configuration of the present embodiment, the same characteristics as in FIG. 38 (e) can be realized. Adder
The output signal of 1108 has a sampling frequency converted from 32.4 MHz to 4 in the sampling frequency conversion circuit 1109.
It is converted to 8.6 MHz and supplied to one of the input terminals of the mixing circuit 112 as a moving area luminance signal processing output.

Next, the moving area color signal processing circuit 1200 is composed of, for example, the circuit shown in FIG. The spectral structures of signals at points a and b in the figure are shown in FIGS. 6 (a) and 6 (b).
Further, the signals D, E, F correspond to the signals at the points D, E, F in FIG. In FIG. 5, the signals D and E are
The time extension circuits 1201 and 1203 respectively set the TCI state of 32.
The color signal area is expanded to the 8.1 MHz rate from the inter-frame interpolated signal of the 4 MHz rate. Time expansion circuit 1201, 12
The output signal of 03 is supplied to vertical low-pass filters 1202 and 1204, respectively.

Here, the output signal of the time expansion circuit 1203 is 1
It is an inter-frame interpolated signal of the signal before the field and the signal before the three fields. Therefore, the vertical low-pass filter 1204 to which it is supplied extracts only the signal one field before from the input signal and performs field interpolation processing using only that signal.

On the other hand, the output signal of the time expansion circuit 1201 is the inter-frame interpolated signal between the current field signal and the signal two fields before, or the inter-frame interpolated signal between the field two fields before and four fields before. It is obtained by police box every time. In the vertical low-pass filter 1202, the current field signal is supplied when the former is supplied, and 2 when the latter is supplied.
Only the signals before the field are extracted, and the field interpolation processing is performed. As a result, both output signals have the spectral structure shown in FIG. Here, the output signal of the vertical low-pass filter 1204 is the field interpolation signal of the 1-field previous signal, and the output signal of the vertical low-pass filter 1202 is the field interpolation signal of the current field / 2-field previous signal. Therefore, it is understood that inter-field processing by frame completion can be realized by performing processing between both.

The field interpolating signals of the 8.1 MHz rate output from the vertical low pass filters 1202 and 1204 are supplied to the inter-field vertical low pass filter 1205. The inter-field vertical low-pass filter 1205 performs inter-field interpolation processing to obtain an output signal having a spectral structure as shown in FIG. 6 (b). The inter-field interpolation processing output having a data rate of 16.2 MHz is supplied to one input terminal of the mixing circuit 113.

The static area luminance signal processing circuit 1300 is composed of, for example, the circuit shown in FIG. Signals G, H, I in the figure are
This corresponds to the signal at each point G, H, I in FIG. In FIG. 7, signals G and H are supplied to sampling frequency conversion circuits 1302 and 1304 after the frequency components of horizontal 12 MHz or less are extracted by the 12 MHz horizontal low-pass filters 1301 and 1303, and the sampling frequency is changed from 32.4 MHz. Two
Converted to 4.3 MHz. Inter-field interpolation processing is performed in the inter-field interpolation circuit 1305 using the output signals from the sampling frequency conversion circuits 1302 and 1304, and the output signal of the sampling frequency conversion circuit 1302 is in the inter-field motion vector correction circuit 1306. It has been motion compensated.

Here, the sampling frequency conversion circuit 1304
Is an interframe interpolated signal of the signal one field before and the signal three fields before. On the other hand, the output signal of the sampling frequency conversion circuit 1302 is an interframe interpolated signal of the current field signal and the signal two fields before, or an interframe interpolated signal of the signal two fields before and the signal four fields before. It is obtained at the police box.
Therefore, if processing is performed between the two, inter-field processing by frame completion can be realized. The inter-field motion vector correction circuit 1306 matches the pixel phase of the inter-frame interpolated signal of the current field and the two-field previous signal, which is the output signal of the sampling frequency conversion circuit 1304, in the same manner as described in the moving area luminance signal processing. Is corrected so that The inter-field motion vector correction circuit 1306 is basically the same as the inter-field motion vector correction circuit in the moving region luminance signal processing shown in FIG.
The delay unit of 52 is 24.3 MHz. The horizontal vector signal has a value indicating the shift amount at the 24.3 MHz rate. Two-dimensional low-pass filter between fields 13
In 05, inter-field interpolation processing is performed by frame completion. The interfield interpolation signal is supplied to the other input terminal of the mixing circuit 112.

The static area color signal processing circuit 1400 is composed of, for example, the circuit shown in FIG. Signals J, K, and L in the figure correspond to signals at points J, K, and L in FIG. FIG.
, The signals J and K are time-expanded four times with respect to the C region of the inter-frame interpolated signal at the rate of 32.4 MHz in the TCI state in the time expansion circuits 1401 and 1402.
Output inter-frame interpolated signal of MHz rate. Here, the output signal of the time expansion circuit 1402 is an interframe interpolated signal of the signal one field before and the signal three field before. On the other hand, the output signal of the time expansion circuit 1401 is an interframe interpolated signal between the current field signal and the signal two fields before, or an interframe interpolated signal between the signal two fields before and the signal four fields before, for each field. It is obtained at the police box. Therefore, if processing is performed between the two, inter-field processing by frame completion can be realized. The output signals of the time expansion circuits 1401 and 1402 are subjected to inter-field interpolation processing by frame completion in an inter-field vertical low pass filter 1403. Inter-field interpolated signal is mixed circuit 11
Supplied to 3.

The processing after the mixing circuit is the same as that described in the conventional example. Wideband R restored as output signal
A GB signal is obtained.

As described above, according to the present embodiment, the field memories 110 and 111 necessary for the interframe interpolation processing, the output signal of the selection circuit 109 and the output signal of the field memory 111 are alternately selected and output for each field. The output signals of the selection circuit 114 and the field memory 1 are
The single-field image signal from 10 is supplied to the moving area luminance signal processing circuit 1100 and the chrominance signal processing circuit 1200. Similarly, the output signal of the selection circuit 115 and the single-field image signal from the field memory 110 are stopped. Area luminance signal processing circuit
By supplying to the 1300 and the color signal processing circuit 1400, it is possible to realize processing equivalent to that of the conventional example, and it is possible to delete the field memory in FIGS. 37, 39, 41 and 42 of the conventional example. it can. (Embodiment 2) FIG. 9 is a block diagram showing the arrangement of a decompression device for band-compressed image signals according to the second embodiment of the present invention.
The difference from the first embodiment shown in FIG. 1 is that the selection circuit 115 in FIG.
The point that the output signal from the selection circuit 114 is supplied to the stationary area color signal processing circuit 1400 instead of the output signal from the selection circuit 115. That is, the moving area processing circuit
The same signal is supplied to 24 and the stationary area processing circuit 25.

In the first embodiment shown in FIG. 1, by separately providing alternate selection circuits for moving field processing and stationary area processing for each field, the phase of frame completion processing can be set separately. ing. However, as an embodiment of high-definition television broadcasting by the image quality improvement type MUSE system, it is assumed that the phase of the frame completion processing is the same in the moving area processing and the still area processing. In such a case, the provision of separate selection circuits is redundant as a circuit configuration. As shown in the second embodiment of the present invention, one selection circuit is provided, and the output signal thereof is the moving area signal processing and the stationary area. It is possible to reduce the circuit scale by using it commonly for signal processing. (Third Embodiment) FIG. 10 is a block diagram showing the arrangement of a band-compressed image signal decompression device according to the third embodiment of the present invention. The difference from the second embodiment shown in FIG. 9 is that the selection circuit
The output signal of 114 is supplied to the signal processing circuit for each of the moving area and the stationary area via the newly provided inter-field vector correction circuit 121, and the moving area luminance signal processing circuit 1100 in FIG. The still area luminance signal processing circuit 1300 replaces the new moving area processing circuit 3100 and the new still area processing circuit 3300.

Although the structure of the moving area luminance signal processing circuit 3100 is not shown, it is the same as the moving area luminance signal processing circuit 1100 shown in FIG. 2 with the inter-field motion vector correction circuit 1110 removed. Similarly, the static area luminance signal processing circuit 33
00 is the same as the stationary area luminance signal processing circuit 1300 shown in FIG. 7 with the inter-field motion vector correction circuit 1306 removed. The inter-field motion vector correction circuit 121 to which the output signal of the selection circuit 114 is supplied has the same configuration as the inter-field motion vector circuit shown in FIG. 4, and the correction direction is linked with the selection phase for each field of the selection circuit 114. It's switching.

As is apparent from FIG. 10, according to the present embodiment, after the inter-field motion vector correction processing is performed on the output signal of the selection circuit 114, the moving area signal processing circuit and the stationary area signal processing circuit are respectively provided. Since it is configured to supply a signal to, the moving area luminance signal processing circuit and the stationary area signal processing circuit each need only one inter-field motion vector correction circuit, and the circuit scale can be reduced. In particular, multi-stage line delay devices are
IC when converted to integrated circuit
The area occupied on the chip is very large, which causes a cost increase, and the merits of commonality are great.

(Embodiment 4) A fourth embodiment of the present invention will be described below with reference to the drawings. FIG. 11 is a block diagram showing the arrangement of a band-compressed image signal decompression device according to the fourth embodiment of the present invention. 11 is different from the configuration shown in FIG. 1 in that it has a second dynamic region luminance signal processing circuit.
The 1500 is newly provided and replaced with the 3-input mixing circuit 116 instead of the mixing circuit 112, and the output signal of the field memory 110 is supplied to the second motion area luminance signal processing circuit.
The input mixing circuit 116 includes a moving area luminance signal processing circuit 110.
0, the output signals from the second moving area luminance signal processing circuit 1500 and the still area luminance signal processing circuit 3300 are supplied.

The second dynamic region luminance signal processing circuit 1500 is composed of, for example, the circuit shown in FIG. Signals A and B in the figure
Correspond to the signals at points A and B in FIG.
In FIG. 12, the signal A is the same as the signal one field before and 3
It is an interframe interpolated signal N1 with the signal before the field. The two-dimensional low-pass filter 1501 extracts only the signal one field before and performs field interpolation processing. The output signal of the two-dimensional low-pass filter 1501 is supplied to the sampling frequency conversion circuit 1502, and the sampling frequency is set to 32.
Converted from 4 MHz to 48.6 MHz. The signal whose sampling frequency has been converted is supplied to the 3-input mixing circuit 116 as a second motion domain processed signal output signal. That is,
In the second moving area luminance signal processing circuit 1500, only intra-field processing is used.

FIG. 13 shows an example of the mixing ratio k (0 ≦ k ≦ 1) with respect to the motion amount in the 3-input mixing circuit 116. The characteristics a, b, and c shown in FIG. 13 are the output signal of the moving region luminance signal processing circuit 1100, the output signal of the second moving region luminance signal processing circuit 1500, and the static state according to the motion detection signal from the control signal input terminal 108. The respective ratios of mixing the output signals of the area luminance signal processing circuit 1300 are shown. The characteristic a is for the moving area luminance signal processing circuit 1100, and the characteristic b is for the second moving area luminance signal processing circuit 15.
The characteristic c corresponds to 00 and the static region luminance signal processing circuit 1300. For example, as shown in the figure, when the movement amount is "x", the mixing ratios are M1, M2, and S, respectively, and the sum thereof is 1.0.

According to this characteristic, the ratio of the static area luminance signal processing circuit 1300 is increased when the static area or the motion amount is small. The ratio of the moving area luminance signal processing circuit 1100 increases as the amount of movement increases, and the ratio of the second moving area luminance signal processing circuit 1500 increases when the amount of movement increases. The image quality deterioration caused when the MUSE signal compressed by the current MUSE band compressing device is restored by the image quality improving MUSE restoring device becomes more remarkable as the motion is faster (the motion amount is larger). According to the decompression device for a band-compressed image signal of the present invention, the faster the motion is, the more the ratio to the output of the second moving region luminance signal processing circuit configured by the intra-field processing is mixed, and thus the faster. Image deterioration due to movement can be reduced.

(Embodiment 5) A fifth embodiment of the present invention will be described below with reference to the drawings. FIG. 14 is a block diagram showing the configuration of a decompression device for band-compressed image signals according to the fifth embodiment of the present invention. 14, the difference from the configuration of FIG. 1 is that the moving region luminance signal processing circuit 1100 of FIG.
However, the third moving area luminance signal processing circuit 1600 is replaced with the third moving area luminance signal processing circuit 1600, and the motion signal applied to the control signal input terminal 108 is supplied to the third moving area luminance signal processing circuit 1600.

The third moving area luminance signal processing circuit 1600 is composed of, for example, the circuit shown in FIG. The spectral structures of signals at points a to e in the figure are shown in FIGS. The signals A, B, and C are the points A, B, and C in FIG.
It corresponds to the signal in. In FIG. 15, signal B
Is an inter-frame interpolated signal N1 of the signal one field before and the signal three fields before. Therefore, the two-dimensional LPF 1604 to which it is supplied extracts only the signal one field before from the signal B and performs field interpolation processing using only that signal. On the other hand, the signal A is an inter-frame interpolated signal N0 between the current field signal and the signal two fields before.
Alternatively, the inter-frame interpolated signal N2 for the field two fields before and the field four fields before are supplied alternately for each field. In the two-dimensional LPF 1601, when the former is supplied, the signal of the current field is extracted, and when the latter is supplied, only the signal of two fields before is extracted and subjected to field interpolation processing. The frequency characteristic of the two-dimensional low pass filter 1601 is shown in FIG. Also, the two-dimensional low-pass filter is 2
A selection signal generation circuit that is a filter having various types of characteristics
The characteristic can be selected according to the selection signal from 1611. Two types of characteristics are shown in FIG.
It shows in (b). The characteristic of FIG. 16B has a narrower band than the characteristic of FIG. Here, the selection signal is "0"
In case of, the characteristic of (b) is selected, and in case of "1", the characteristic of (a) is selected. On the other hand, the control signal input terminal 108
The motion detection signal supplied from the
Is supplied to. The selection signal generation circuit 1611 is shown in FIG.
In the circuit having the conversion characteristics as shown in, up to a preset value of the input motion detection signal,
"1" is output as the selection signal, and "0" is output when it exceeds the selection signal. The selection signal is supplied to the two-dimensional low-pass filter 1611 and the selection circuit 1609.

First, the case where the motion detection signal is small (the selection signal is "1") will be described. In the two-dimensional low-pass filters 1601 and 1604, field interpolation processing by the filter having the characteristic of FIG.
An output signal having a spectral structure as shown in (a) is obtained. The output signal of the two-dimensional low-pass filter 1601 is subjected to motion vector correction by the inter-field motion vector correction circuit 1612 and supplied to the 8 MHz horizontal low-pass filter 1602 and the subtractor 1603. Here, the inter-field motion vector correction circuit 1612
Is the same as that shown in FIG. The output signal of the two-dimensional low-pass filter 1604 is the 8 MHz horizontal low-pass filter 1605.
And to the subtractor 1606. 8MHz horizontal low-pass filter
In 1602 and 1605, horizontal frequency components of 8 MHz or less are extracted, and both become output signals having a spectral structure as shown in FIG. 18 (b). The subtractors 1603 and 1606 perform difference processing between the output signal of the two-dimensional low-pass filter and the output signal of the 8 MHz horizontal low-pass filter to obtain a high-pass signal of 8 MHz or more in the horizontal direction, both of which are shown in FIG. The output signal has such a spectral structure. Output signals of the subtractors 1603 and 1606 are supplied to an inter-field vertical low pass filter 1607. The vertical low-pass filter between fields 1607 is horizontal 8M
A high-frequency signal of Hz or higher is subjected to vertical low-pass filter processing between fields, and an output signal having a spectral structure as shown in FIG. 18D is obtained. In the adder 1608, addition processing of the output signal of the 8 MHz horizontal low-pass filter 1605 and the output signal of the inter-field vertical low-pass filter 1607 is performed.
The output signal has a spectrum structure as shown in (e) and is supplied to one input terminal of the selection circuit 1609. A two-dimensional low-pass filter 1604 is connected to the other input terminal of the selection circuit 1609.
Are supplied. The selection circuit 1609 selects the output signal from the adder 1608 when the selection signal is "1" according to the selection signal from the selection signal generation circuit 1611, and the two-dimensional low range when the selection signal is "0". It operates to select the output signal from the filter 1604. Therefore, the output signal from the adder 1608 is selected as the output signal of the selection circuit 1609, the sampling frequency is converted from 32.4 MHz to 48.6 MHz in the sampling frequency conversion circuit 1610, and the third moving region luminance signal processing circuit It is supplied to the mixing circuit 112 as an output signal. As described above, when the motion detection signal is small, the same operation as the moving region luminance signal processing in the image quality improvement type MUSE restoration apparatus as described in the first embodiment is realized.

Next, the case where the motion detection signal is large (the selection signal is "0") will be described. Select signal is "0"
In the case of, the two-dimensional low-pass filter 1604 performs field interpolation processing using a filter having a narrow band characteristic as shown in FIG. 16B, and the field interpolation signal is the selection circuit 1609.
The sampling frequency is converted from 32.4 MHz to 48.6 MHz in the sampling frequency conversion circuit 1610 via the, and is supplied to the mixing circuit 112 as an output signal of the third moving region luminance signal processing circuit. When the detection signal is large in this way, only the processing within the field is switched and the characteristic is set to a narrow band.

As described above, according to this embodiment, the selection signal is generated according to the magnitude of the motion detection signal, and when there is a large movement, the two-dimensional low-pass filter 1604 is selected by the selection circuit 1609 and the field Because the inter-process is passed through,
Similar to the one described in the embodiment, it is possible to prevent the image quality deterioration due to a large movement. Further, since the two-dimensional low-pass filter having a narrower band characteristic is selected, it is possible to reduce the occurrence of aliasing interference due to not using the inter-field processing.

(Embodiment 6) A sixth embodiment of the present invention will be described below with reference to the drawings. FIG. 19 is a block diagram showing the configuration of a band-compressed image signal decompression device according to the sixth embodiment of the present invention. 19, the difference from the configuration shown in FIG. 1 is that the first median value extraction circuit 1800 and the second median value extraction circuit 1900 are connected in series to the output end of the moving region color signal processing circuit 1200, The output terminal of the second median value extraction circuit 1900 is connected to the mixing circuit 113.

When the MUSE signal compressed by the current MUSE band compressing device is restored by the image quality improving type MUSE restoring device, multi-line interference occurs first in the edge part of the high-saturation image with fast movement. As stated. This is an inter-field interpolation process (in the figure, an inter-field vertical low-pass filter 12 in the moving-region color signal processing circuit shown in FIG. 5).
Equivalent to 05). Therefore, the generated multi-line interference is 8.1M as shown in FIG.
It is a high frequency component of Hz.

For example, it is assumed that such a disturbed signal is output from the moving area color signal processing circuit 1200.
The output signal is supplied to the first median value extraction circuit 1800. The output signal of the first median value extraction circuit 1800 is supplied to the second median value extraction circuit 1900. First median extraction circuit 1800
The second median value extraction circuit 1900 is composed of, for example, the circuit shown in FIG. Waveforms of signals at points A to C in the figure are shown in FIGS. The input signal is delayed pixel by pixel by the one-pixel delay units 1801 and 1802 (the delay amount of the one-pixel delay unit is 1 / 16.2 MHz), and the delayed signal is supplied to the median value determination circuit 1803. Median value determination circuit
In 1803, the levels of the signals for the supplied three pixels are compared, and the second level is output. For example, the pixel of interest
If “b”, then the median discriminating circuit 1803 is
The sizes of “a”, “b”, and “c” are compared. The median value in this case is a level corresponding to the pixel “a” or “c”. Is Figure 2
The level indicated by the pixel "e" of 1 (B) is output. With the above operation, the waveform shown in FIG. 21B is obtained from the first median value extraction circuit 1800 and supplied to the median value extraction circuit 1900. The median value extraction circuit 1900 has the same configuration as the first median value extraction circuit 1800, and as a result, FIG.
The waveform shown in is obtained and multi-line interference can be eliminated.

As described above, according to this embodiment, the moving area color signal processing circuit 1200 includes the first median value extracting means 1800 and the second median value extracting means 1800 for extracting the median level from the pixels in the predetermined area. By installing the extraction means 1900 in series, the current MUS
Image quality improvement type MUSE for MUSE signal compressed by E method
There is a possibility that it will occur when you restore with a system restorer,
It is possible to reduce multi-line interference at an edge portion of a high-saturation image that moves fast.

(Embodiment 7) A seventh embodiment of the present invention will be described below with reference to the drawings. FIG. 22 is a block diagram showing the arrangement of a band-compressed image signal decompression device according to the seventh embodiment of the present invention. This embodiment is constructed for the same purpose as that of the sixth embodiment shown in FIG. The difference from the configuration of the sixth embodiment shown in FIG. 19 is that the first median value extraction circuit 1800 and the second median value extraction circuit 1900 connected in series are the maximum value extraction circuit 2000 and the minimum value extraction circuit. It was replaced by 2100 respectively.
The maximum value and minimum value extraction circuits 2000 and 2100 are shown in FIG.
The circuit shown in FIG. The signal waveforms at points A to C in the figure are shown in FIGS. In FIG. 23, the input signal A is 8. A as shown in FIG.
It is assumed that the signal has a multi-line interference component of 1 MHz.
The input signal is delayed by one pixel by the one-pixel delay units 2001 and 2002, and the delayed signal is supplied to the maximum value determination circuit 2003. The maximum value discriminating circuit 2003 discriminates and outputs the largest level from the supplied signals for three pixels.
For example, when the pixel of interest is “B”, the maximum value determination circuit
In 2003, the sizes of pixels “a”, “b”, and “c” are compared. The maximum value in this case is the level corresponding to the pixel “C”. Therefore, for the pixel of interest "B",
The level shown in the pixel "e" of 4 (B) is output. As a result of the above-described operation, the maximum value extraction circuit 2000 is operated as shown in FIG.
The waveform shown in (B) is obtained and supplied to the minimum value extraction circuit 2100.

The input signal to the minimum value extraction circuit 2100 is delayed by one pixel by the 1-pixel delay units 2101, 2102, and the delayed signal is supplied to the minimum value determination circuit 2103. The minimum value discriminating circuit 2103 discriminates and outputs the smallest level from the supplied signals for three pixels. For example, when the pixel of interest is “e”, the pixel is
The sizes of “d”, “e”, and “f” are compared. The minimum value in this case is the level corresponding to the pixel “d”. The level shown in the pixel "C" of C) is output.By the above operation, the waveform shown in Fig. 24C is obtained from the minimum value extraction circuit 2100, and the multi-line interference can be removed.

In this embodiment, the maximum value extraction circuit
Although the output signal of 2000 is supplied to the minimum value extraction circuit 2100, the present invention is not limited to this, and the connection order may be changed.

(Embodiment 8) An eighth embodiment of the present invention will be described below with reference to the drawings. FIG. 25 is a block diagram showing the structure of the moving area luminance signal processing circuit in the eighth embodiment of the present invention. In the moving region luminance signal processing circuit 1700 of FIG. 25, the input signals A and B are two-dimensional low pass filters.
They are connected to 1701 and 1702 respectively. The output terminal of the two-dimensional low-pass filter 1701 is connected to one input terminal of the inter-field vertical low-pass filter 1703 via the inter-field motion vector correction circuit. Two-dimensional low-pass filter 17
The output terminal of 02 is connected to the other input terminal of the inter-field vertical low-pass filter 1703 and one input terminal of the subtractor 1704,
The output terminal of the inter-field vertical low-pass filter 1703 is a subtractor 17
It is connected to the other input terminal of 04 and one input terminal of the adder 1706. The output terminal of the subtracter 1704 is connected to the other input terminal of the adder 1706 via an 8 MHz horizontal low-pass filter 1705. The output terminal of the adder 1706 is connected to the sampling frequency conversion circuit 1707, and the sampling frequency conversion circuit 17
The signal C is obtained from the output terminal of 07.

Regarding the operation of the moving area luminance signal processing circuit in the eighth embodiment configured as described above, the moving area luminance signal processing circuit 1100 in the restoring apparatus shown in FIG.
It is assumed that the moving area luminance signal processing circuit 1700 of FIG. Except for the operation in the moving area luminance signal processing circuit 1700, it is the same as that described in the first embodiment. Figure 25
In the dynamic region luminance signal processing circuit 1700 shown in FIG. 26, the spectral structure of the signal at each point a to e in the figure is shown in FIG.
~ (E). The signals A, B and C correspond to the signals at points A, B and C in FIG.

In FIG. 25, the input signal B is an interframe interpolated signal of the signal one field before and the signal three field before. Therefore, in the two-dimensional LPF1702, the signal B
From, only the signal one field before is extracted and subjected to field interpolation processing. On the other hand, the input signal A is an interframe interpolated signal of the current field signal and a signal two fields before, or an interframe interpolated signal of the signal two fields before and four fields before is alternately supplied for each field. There is. In the two-dimensional LPF 1701, when the former is supplied, the signal of the current field is extracted, and when the latter is supplied, only the signal of two fields before is extracted and subjected to field interpolation processing. As a result, both output signals have the spectral structure shown in FIG.

The output signal of the two-dimensional LPF 1101 is inter-field motion vector corrected in the inter-field motion vector correction circuit 1708. The inter-field motion vector correction circuit 1708 is, for example, the same as that shown in FIG.
The correction direction is switched in synchronism with the input signal A that alternates for each field.

The inter-field vertical interpolation low-pass filter 1703 outputs inter-field interpolated signals output from the inter-field motion vector correction circuit 1708 and the two-dimensional low-pass filter 1702.
The output signal has a spectral structure as shown in 6 (b). The subtractor 1704 subtracts the output signal of the two-dimensional low-pass filter 1702 from the output signal of the inter-field vertical low-pass filter 1703 to obtain an output signal having a spectrum structure as shown in FIG. 17 (c). The 8 MHz horizontal low-pass filter 1705 extracts frequency components below 8 MHz in the horizontal direction from the signal having the spectral structure shown in FIG.
The output signal has a spectral structure as shown in (d). In the adder 1706, the inter-field vertical low pass filter 17
The output signal of 03 and the output signal of the 8 MHz horizontal low-pass filter 1705 are added, and an output signal having a spectrum structure as shown in FIG. 17E is obtained. Therefore, it can be understood that the processing according to the present embodiment can realize the same characteristics as those of the moving area luminance signal processing circuit shown in FIG. The output signal of the adder 1706 is supplied to the sampling frequency conversion circuit 1707 and the sampling frequency is changed from 32.4 MHz to 48.6.
Converted to MHz and output as signal C. The signal C is supplied to the mixing circuit 112 of FIG.

As described above, according to this embodiment, the inter-field vertical low-pass filter 1703 is used for the two-dimensional low-pass filter 170.
Inter-field processing is performed using the image signals of the respective fields from 1 and 1702, and the subtractor 1704 subtracts the output signal of the two-dimensional low-pass filter 1702 from the output signal of the inter-field vertical low-pass filter 1703. The 8 MHz horizontal low-pass filter 1705 extracts the horizontal low-pass frequency component from the output signal of the subtractor 1704, and the adder 1706 outputs the output signal of the inter-field vertical low-pass filter 1703 and the 8 MHz horizontal low-pass filter. By adding the output signals of 1705,
Two 8M used in the moving area luminance signal processing circuit 1100 of
The Hz horizontal low-pass filters 1102 and 1105 can be reduced to one.

In the description of the present embodiment, the case where the present invention is applied to the decompression device for the band-compressed image signal shown in FIG. 1 has been described, but the present invention is not limited to this, and FIGS.
It can also be applied to the configurations shown in FIGS. FIG.
When applied to the configuration shown in FIG. 2, since the input signal A has already been inter-field motion vector corrected, FIG.
5, the inter-field motion vector correction circuit 1708 is unnecessary.

[0088]

As described above, the first and second
According to the invention, the image quality improvement type MUSE restoration apparatus can be realized without increasing the memory capacity as compared with the current MUSE restoration apparatus.

Further, according to the third and fourth inventions, the circuit scale can be reduced.

Further, according to the fifth, sixth, seventh, and eighth inventions, the compatibility between the current MUSE method and the image quality improving MUSE method, particularly, the current MUSE method is used.
Image quality improvement type M
It is possible to reduce interference that occurs when data is restored by the USE restoration device.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a band-compressed image signal decompression device according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of a moving area luminance signal processing circuit in the embodiment.

FIG. 3 is an operation spectrum structure diagram of a moving region luminance signal processing circuit in the embodiment.

FIG. 4 is a block diagram showing a configuration of a motion vector correction circuit in the same embodiment.

FIG. 5 is a block diagram showing a configuration of a moving area color signal processing circuit in the embodiment.

FIG. 6 is an operation spectrum structure diagram of the moving region color signal processing circuit in the embodiment.

FIG. 7 is a block diagram showing the configuration of a static area luminance signal processing circuit in the embodiment.

FIG. 8 is a block diagram showing a configuration of a still area color signal processing circuit in the embodiment.

FIG. 9 is a block diagram showing the configuration of a decompression device for band-compressed image signals according to a second embodiment of the present invention.

FIG. 10 is a block diagram showing the configuration of a band-compressed image signal decompression device according to a third embodiment of the present invention.

FIG. 11 is a block diagram showing a configuration of a band-compressed image signal decompression device according to a fourth embodiment of the present invention.

FIG. 12 is a block diagram showing a configuration of a second moving region luminance signal processing circuit in the same embodiment.

FIG. 13 is a characteristic diagram for explaining the operation of the 3-input mixing circuit 116 in the embodiment.

FIG. 14 is a block diagram showing the configuration of a decompression device for band-compressed image signals according to a fifth embodiment of the present invention.

FIG. 15 is a block diagram showing a configuration of a third moving area luminance signal processing circuit in the embodiment.

FIG. 16 is an operating frequency characteristic diagram of the two-dimensional low-pass filter in the example.

FIG. 17 is a conversion characteristic diagram of the selection signal generation circuit 1611 in the same embodiment.

FIG. 18 is an operational spectrum structure diagram of the moving region luminance signal processing circuit in the embodiment.

FIG. 19 is a block diagram showing the configuration of a band-compressed image signal decompression device according to a sixth embodiment of the present invention.

FIG. 20 is a block diagram showing the configuration of a median value extraction circuit in the embodiment.

FIG. 21 is a waveform diagram for explaining the operation of the median extraction circuit in the same example.

FIG. 22 is a block diagram showing a configuration of a band-compressed image signal decompression device according to a seventh embodiment of the present invention.

FIG. 23 is a block diagram showing the configurations of a maximum value extraction circuit and a minimum value extraction circuit in the same embodiment.

FIG. 24 is a waveform chart for explaining the operation of the maximum value extraction circuit and the minimum value extraction circuit in the same embodiment.

FIG. 25 is a block diagram showing a configuration of a moving area luminance signal processing circuit according to an eighth embodiment of the present invention.

FIG. 26 is an operation spectrum structure diagram of the moving region luminance signal processing circuit in the embodiment.

FIG. 27 is a block diagram showing the configuration of an image quality improvement type MUSE band compression device.

FIG. 28 is a block diagram showing a configuration of a moving region luminance signal processing circuit of the compression device.

FIG. 29 is an operation spectrum structure diagram of a moving region luminance signal processing circuit of the compression device.

FIG. 30 is a block diagram showing a configuration of a static area luminance signal processing circuit of the compression apparatus.

FIG. 31 is a block diagram showing a configuration of a color signal processing circuit of the compression device.

FIG. 32 is an operational spectrum structure diagram of the color signal processing circuit of the compression device.

FIG. 33 is a block diagram showing a configuration of a moving region color signal processing circuit of the compression device.

FIG. 34 is an operational spectrum structure diagram of the moving region color signal processing circuit of the compression device.

FIG. 35 is a block diagram showing a configuration of a still region color signal processing circuit of the compression device.

FIG. 36 is a block diagram showing the configuration of a decompression device for a band-compressed image signal of an image quality improvement type MUSE method.

FIG. 37 is a block diagram showing a configuration of a moving area luminance signal processing circuit of the restoration device.

FIG. 38 is an operation spectrum structure diagram of the moving region luminance signal processing circuit of the restoration device.

FIG. 39 is a block diagram showing a configuration of a moving region color signal processing circuit of the restoration device.

FIG. 40 is an operational spectrum structure diagram of the moving region color signal processing circuit of the restoration device.

FIG. 41 is a block diagram showing a configuration of a still region luminance signal processing circuit of the restoration device.

FIG. 42 is a block diagram showing a configuration of a still region color signal processing circuit of the restoration device.

[Explanation of symbols]

1 Input terminal 2 Output terminal 5 Inverse matrix circuit 6 Line sequential decoding circuit 13 Interframe interpolation circuit 11 AD converter 21 DA converter 20 Interfield processing circuit 24 Motion area processing circuit 25 Static area processing circuit 106, 107, 108 Control Signal input terminal 109, 114, 115, 1609 Selection circuit 110, 111 Field memory 112, 113, 116 Mixing circuit 1100, 1500, 1600, 1700 Motion area luminance signal processing circuit 1101, 1104, 1501, 1601, 1604, 1701, 1702 Two-dimensional low-pass filter 1102, 1105, 1602, 1605, 1705 8MHz horizontal low-pass filter 1103, 1106, 1603, 1606, 1704 Subtractor 1107, 1205, 1403, 1607, 1703 Vertical low-pass filter between fields 1108, 1608, 1706 Adder 1109, 1302, 1304, 1502, 1610, 1707 Sampling frequency conversion circuit 1200 Motion area color signal processing circuit 1201, 1203, 1401, 1402 Time expansion circuit 1202, 1204 Vertical low pass filter 1300 Still area luminance signal processing circuit 1301, 1303 12M z horizontal low pass filter 1305 field between the two-dimensional low-pass filter 1400 still region color signal processing circuit 1800,1900 median extraction circuit 1801,1802,1901,1902,2001,2002,2101,2102 1
Pixel delay device 1803, 1903 Median value determination circuit 2000 Maximum value extraction circuit 2003 Maximum value determination circuit 2100 Minimum value extraction circuit 2103 Minimum value determination circuit

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Yoichiro Miki 1006 Kadoma, Kadoma City, Osaka Prefecture Matsushita Electric Industrial Co., Ltd. (72) Masaki Tokoi 1006 Kadoma, Kadoma City, Osaka Matsushita Electric Industrial Co., Ltd. 72) Inventor Kazuya Ueda 1006 Kadoma, Kadoma-shi, Osaka Prefecture Matsushita Electric Industrial Co., Ltd. (72) Inventor Masanori Hamada Osaka 1006 Kadoma, Kadoma Matsushita Electric Industrial Co., Ltd. 1006, Kadoma, Kadoma, Fuchu, Matsushita Electric Industrial Co., Ltd. (72) Inventor Yuichi Ninomiya 1-10-11 Kinuta, Setagaya-ku, Tokyo Inside the Japan Broadcasting Corporation Broadcasting Technology Laboratory (72) Inventor Seiichi Koshi Setagaya, Tokyo 1-10-11 Kukina, Japan Broadcasting Corporation Research Institute for Broadcasting Systems (72) Inventor Yoshinori Izumi Kinichi Setagaya-ku, Tokyo No. 10-11 Broadcasting Technology Institute of Japan Broadcasting Corporation (72) Inventor Masahide Naemura 1-10-11 Kinuta, Setagaya-ku, Tokyo Broadcasting Technology Research Institute of Japan Broadcasting Association (72) Inventor Koichi Yamaguchi Setagaya-ku, Tokyo 1-10-11 Kinuta, Broadcasting Technical Research Institute of Japan Broadcasting Corporation (72) Inventor Atsushi Fukuda 1-10-11 Kinuta, Setagaya-ku, Tokyo Broadcasting Engineering Research Institute of Japan Broadcasting Corporation

Claims (18)

[Claims] 1. A decompression device for a high-definition television signal band-compressed using a multiple sub-sampling method, wherein a video signal interpolated between frames using the band-compressed high-definition television signal as an input is provided. From the delay means that has been sent out and the three video signals of N0, N1 and N2 which are temporally continuous and which can be obtained from this delay means,
And a signal selecting means for selecting N2 for each field, and the inter-field processing is performed using the video signal output by the selecting means and the video signal of N1. Restoration device. 2. A moving area processing means for inputting a high-definition television signal band-compressed using the multiple sub-sampling method and performing different interpolation processing for the moving area and the still area, and a still area processing means. A decompression device for decompressing the original high-definition television signal by weight-mixing the respective output signals according to the motion of the image, wherein the inter-frame interpolation processing is performed on the band compression signal. An output signal from the interpolating means and the interframe interpolating means,
Field delay means for extracting inter-frame interpolated signals for three consecutive fields of 0, N1 and N2, first selection means for selectively outputting one of N0 and N2, and N0 and N2 A second selecting means for selectively outputting either one, supplying the N1 and the output signal of the first selecting means to the moving area processing means, and outputting the N1 and the second selecting means. And a signal to the stationary area processing means,
The band-compressed image signal decompression device is characterized in that the selection means and the second selection means switch the selection output for each field. 3. A moving area processing means for inputting a high-definition television signal band-compressed using the multiple sub-sampling method and performing different interpolation processing for the moving area and the still area, and a still area processing means. A decompression device for decompressing the original high-definition television signal by weight-mixing the respective output signals according to the motion of the image, wherein the inter-frame interpolation processing is performed on the band compression signal. An output signal from the interpolating means and the interframe interpolating means,
Field delay means for extracting inter-frame interpolated signals for three consecutive fields of 0, N1 and N2, and selection means for selectively outputting either one of N0 and N2 are provided. An apparatus for decompressing a band-compressed image signal, wherein an output signal is supplied to the moving area processing means and the still area processing means, and the selecting means switches the selected output for each field. 4. A moving area in which a high-definition television signal, which is band-compressed using the multiple sub-sampling method and contains motion vector information for motion correction, is input, and different interpolation processing is applied to the moving area and the still area. A decompression device that includes a processing unit and a stationary region processing unit, and that restores the original high-definition television signal by weight-mixing the respective output signals according to the movement of the image, wherein An interframe interpolation means for performing interframe interpolation processing while performing motion correction based on motion vector information, and an output signal from the interframe interpolation means are input, and N
Field delay means for extracting inter-frame interpolated signals for three consecutive fields of 0, N1 and N2, selection means for selectively outputting one of N0 and N2, and an output signal from the selection means Motion vector correction means for performing motion correction based on motion vector information, and supplying the N1 and the output signal from the motion vector correction means to the motion area processing means and the stationary area processing means, The means switches the selective output for each field, and the motion vector correcting means performs the motion correction so that the output signal from the selecting means is aligned with the image phase of the N1. Restoration device. 5. The moving region processing means includes first field interpolation means for performing field interpolation processing on the N0 or N2 interframe interpolation signal, and the N1 interframe interpolation signal. Second field interpolation means for performing field interpolation processing, and first high frequency component extraction from the field interpolation signal output from the first field interpolation means
Second horizontal high frequency component extraction means for extracting horizontal high frequency component from the field interpolation signal output from the second high frequency component extraction means and the second field interpolation means; Horizontal low frequency component extraction means for extracting a horizontal low frequency component from the field interpolation signal output from the second field interpolation means; and a horizontal extracted from the first horizontal high frequency component extraction means. A vertical low frequency component extracting means for extracting a vertical low frequency component by inter-field calculation using a high frequency component and a high frequency component extracted from the second horizontal high frequency component extracting means; A vertical low frequency component extracted from the vertical low frequency component extracting means and an adding means for adding the horizontal low frequency component extracted from the horizontal low frequency component extracting means; Claim 2, characterized in Rukoto claim 3, or claim 4 restoration device bandwidth compression image signals according. 6. The moving area processing means includes a first field interpolation means for performing field interpolation processing on the N0 or N2 interframe interpolation signal, and the N1 interframe interpolation signal. Second field interpolation means for performing field interpolation processing, field interpolation signals output from the first field interpolation means, and output from the second field interpolation means. A vertical low frequency component extracting means for extracting a vertical low frequency component by an inter-field operation using the field interpolating signal, and a field interpolating signal output from the first field interpolating means. Vertical high frequency component extraction means for extracting a vertical high frequency component by inter-field calculation using the field interpolation signal output from the second field interpolation means; A horizontal low-frequency component extracting means for extracting a horizontal low-frequency component from the vertical high-frequency component extracted from the vertical high-frequency component, and a vertical extracted from the vertical low-frequency component extracting means. The low frequency component and the adding means for adding the horizontal low frequency component extracted from the horizontal low frequency component extracting means are provided. Decompressor for band-compressed image signal. 7. A second moving area processing means for performing moving area processing only by in-field processing is provided separately from the moving area processing means, and an output signal of the moving area processing means and the second moving area processing means are provided. 5. The band-compressed image according to claim 2, wherein the output signal of the area processing means and the output signal of the stationary area signal processing means are weight-mixed according to the movement of the image. Signal restoration device. 8. The moving area processing means is replaced with a third moving area processing means for switching between intra-field processing and inter-field processing according to the movement of an image. Alternatively, the decompression device for the band-compressed image signal according to claim 4. 9. A third motion area processing means comprises: first field interpolation means for performing field interpolation processing on the N0 or N2 interframe interpolation signal; Second field interpolation means for subjecting the interpolation signal to field interpolation processing; and a first high frequency component extracted from the field interpolation signal output from the first field interpolation means. 1
Second horizontal high frequency component extraction means for extracting horizontal high frequency component from the field interpolation signal output from the second high frequency component extraction means and the second field interpolation means; Horizontal low frequency component extraction means for extracting a horizontal low frequency component from the field interpolation signal output from the second field interpolation means; and a horizontal extracted from the first horizontal high frequency component extraction means. A vertical low frequency component extracting means for extracting a vertical low frequency component by inter-field calculation using a high frequency component and a high frequency component extracted from the second horizontal high frequency component extracting means; Adding means for adding the vertical low-frequency component extracted by the vertical low-frequency component extracting means and the horizontal low-frequency component extracted by the horizontal low-frequency component extracting means; A selector is provided for selectively outputting the output signal from the adder and the output signal of the second field interpolator, and the selector is the second field interpolator when the motion of the image is large. 9. The decompression device for a band-compressed image signal according to claim 8, wherein the decompression device is controlled so as to select the output signal. 10. The second field interpolation means has a wide band two-dimensional low-pass characteristic and a narrow band two-dimensional low-pass characteristic, and is a narrow band two-dimensional when the image movement is large. The decompression device for a band-compressed image signal according to claim 9, wherein the decompression device switches to a low-pass characteristic. 11. The moving area processing means includes moving image color signal processing means, and the moving image color signal processing means performs field interpolation processing on the N0 or N2 interframe interpolation signals. Output from the first field interpolation means; second field interpolation means for performing field interpolation processing on the N1 interframe interpolation signal; And a field interpolating means for performing interfield interpolating processing using the field interpolating signal and the field interpolating signal output from the second field interpolating means. The band-compressed image signal decompression device according to claim 2, claim 3, or claim 4. 12. A plurality of stages of median value extracting means for extracting the median level from pixels in a predetermined area are provided in series at the output end of the moving area color signal processing means.
A decompression device for the band-compressed image signal described. 13. A median value extracting means comprises a delay means for delaying a pixel and a median value judging means for judging a median value when a plurality of pixel signals from the delay means are inputted and arranged in descending order of level. 13. The decompression device for a band-compressed image signal according to claim 12, further comprising: 14. A maximum value extracting means for extracting a maximum value from pixels in a predetermined area at an output end of the moving area color signal processing means,
12. The decompression device for a band-compressed image signal according to claim 11, further comprising: a minimum value extracting means for extracting a minimum value from pixels in a predetermined area. 15. The maximum value extraction means comprises a delay means for delaying a pixel, and a maximum value determination means for inputting a plurality of pixel signals from the delay means and determining the maximum value of the level. 15. The decompression device for a band-compressed image signal according to claim 14. 16. The minimum value extraction means comprises a delay means for delaying a pixel and a minimum value determination means for inputting a plurality of pixel signals from the delay means and determining the smallest value of the level. 15. The decompression device for a band-compressed image signal according to claim 14. 17. The still region processing means includes a first horizontal low frequency component extracting means for extracting a horizontal low frequency component from the N0 or N2 inter-frame interpolated signal, and the N1 inter frame interpolated signal. Second horizontal low-frequency component extracting means for extracting a horizontal low-frequency component from the first horizontal low-frequency component, and inter-field interpolation processing using the first horizontal low-frequency component and the second horizontal low-frequency component The band compression image signal decompression device according to claim 2, 3 or 4, further comprising: inter-field interpolation means. 18. The still area processing means comprises still area color signal processing means, wherein the still area color signal processing means comprises the N0 or N2 inter-frame interpolated signal and the N1 inter-frame interpolated signal. The decompression device for a band-compressed image signal according to claim 2, 3 or 4, further comprising: inter-field interpolation processing means for performing inter-field interpolation processing using the interpolated signal.