JP2905234B2 - Image signal band compression transmission device and band compression image signal reception device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to band compression transmission of television signals.
In particular, it relates to motion adaptive sub-sampling transmission of high definition television signals.

SUMMARY OF THE INVENTION The present invention relates to band compression transmission of a high definition television signal. In transmitting a high definition television signal after performing motion adaptive subsample band compression, the subsampling is performed by a field offset / frame inversion subsampling. At the same time, prior to this sampling, a predetermined part of a relatively high frequency region of the still image two-dimensional spectrum is moved to another predetermined spectral region.

By doing so, it is possible for the receiving side to restore about 3/4 of the information of the still image with about 1/2 of the transmission information due to the field offset, and also to easily detect the moving area.

The (prior art) system for transmitting a wideband high definition television signal band-compressed, MUSE made in the development of this application's (Mu ltiple Sub
-Nyquist Sampling Encoding) method, and the details are described in the literature, Ninomiya et al .: Development of MUSE method, NHK technical research, Vol.39, No.2, pp.18-53 (1987). It is the source of technology development and deformation development.

The MUSE method described in the above document is a motion adaptive and motion compensation type field offset, frame offset multiplex sub-sample band compression method in which a region where aliasing distortion is not mixed below 4 MHz is provided, and motion detection by inter-frame signals is performed in this band. ing. This method is called a MUSE-E method, and sampling of a still image is performed in four fields.

In contrast to this method, there is a motion adaptive field offset sub-sampling band compression method in which the degree of band compression is reduced to about 1/2, but still image sampling is performed in two fields.
This is called the MUSE-T system, and can maintain high image quality as compared with the MUSE-E system.

(SUMMARY invention) shown in doing simple field offset sub-sampling by the foregoing MUSE-T method (the sampling pattern obtained by the field offset sampling Figure 10 (a). Incidentally, x ₀ , y ₀ is the interval between the original sampling points in the horizontal and vertical directions and is expressed as 1 / frequency.) In this sampling method, however, the arrangement of the sampling carriers generated by the sampling pattern is determined by the time frequency T = 0 Hz. In FIG. 10 (b), a grid formed by sampling carriers corresponding to horizontal and vertical primitive sample periods (in FIG. 10 (b),
X and Y axes and coordinates (1 / x ₀ , 0), (−1 / x ₀ , 0), (0,1 /
_{y 0), (0, -1} / y 0) through to the carrier (FIG. 10 caused by subsampling the center of the grating) formed by the straight line parallel to the X axis · Y-axis (b), two Because of the presence of a double circle）), the frequency domain of the still image that can be transmitted because of the need to avoid aliasing distortion is half (area ratio) of the Nyquist frequency domain corresponding to the original sampling without carriers caused by sub-sampling. ), In the MUSE-T system, the transmission band is band-limited to a triangle in the horizontal and vertical two-dimensional spectrum planes. There is a problem that the vertical resolution of a high portion is insufficient and the horizontal resolution of a portion having a high vertical frequency is insufficient.

In order to avoid this, if a sampling pattern of the field offset / frame inversion subsampling of the four-field sequence is used (the sampling pattern obtained by the field offset / frame inversion subsampling is shown in FIG. 1), it is generated by the sampling pattern. FIGS. 2 (a) to 2 (d) show the arrangement of sampling carriers on a three-dimensional frequency domain composed of horizontal, vertical and time frequency axes. In particular, as shown in FIG. 4 on the surface of T = 15Hz
Since sampling carriers are generated due to the field sequence, it is necessary to perform motion detection on a plane (not shown) with a time frequency T = 7.5 Hz. this is,
This means that motion is detected between two frames, and erroneous motion detection is increased.

Therefore, an object of the present invention is to eliminate the drawbacks of the conventional simple field offset sub-sampling, expand the transmittable area of the still image area, and at the same time, implement an image signal band compression transmitting apparatus and a band which can easily detect image motion. It is intended to provide a compressed image signal receiving device.

(Means for Solving the Problems) In order to achieve this object, an image signal band compression transmission apparatus according to the present invention is an image signal band compression transmission apparatus for transmitting a high-definition television signal after performing motion adaptive sub-sample band compression. Band control means for removing horizontal and vertical high frequency components of a still image transmittable area by primitive sampling;
Sampling means for compressing a sub-sample band by field offset / frame inversion subsampling in which a sampling pattern makes a round in four fields, and a still image 2 from which the horizontal and vertical high frequency components have been deleted prior to the field offset / frame inversion subsampling A region corresponding to a region near the zero frequency of the horizontal and vertical spatial frequencies in the still image two-dimensional spectrum in which the time frequency after the field offset / frame inversion subsampling is a frame frequency / 2, which is a partial region of the two-dimensional spectrum. The horizontal and time frequencies are both zero and the vertical frequency is the original sampling vertical frequency / 4. And a spectrum moving means for moving the spectrum to the vertical high frequency side of the vertical high frequency range.

The band-compressed image signal receiving apparatus of the present invention is a band-compressed image signal receiving apparatus for receiving a high-definition television signal transmitted by the image-signal band-compressing transmitting apparatus of the present invention. Interpolating means for generating a still image two-dimensional spectrum signal from which aliasing components have been removed, and a spectrum shift for moving a part of the still image two-dimensional spectrum, which has been moved in advance at the time of transmission, back to the original area And at least means.

(Operation) According to the present invention, in addition to the conventional field offset sub-sampling, the sampling pattern is inverted for each frame, so that the sub-sampling carrier is
Offset in the time frequency direction of 15 Hz, the time frequency T
The area in which a still image can be transmitted in the 0 Hz plane is enlarged. Further, in this state, the motion detection is performed at the time frequency T =
Although it must be performed on the 7.5 Hz plane, by removing a part of the spectral region before performing sub-sampling, the part of the spectrum that folds back to the plane of time frequency T = 15 Hz is eliminated, By leaving a window for motion detection, motion detection in a plane with a time frequency T = 15 Hz, that is, motion detection based on a frame difference is enabled.

(Examples) Hereinafter, the present invention will be described in detail by examples with reference to the accompanying drawings.

FIG. 1 shows a sampling pattern of the field offset / frame inversion subsampling according to the present invention, and FIG. 2 shows an arrangement example corresponding to the sampling pattern of the sampling carrier.

In FIG. 1, the circles, squares, triangles, and crosses indicate sampling positions of luminance signals in the first, second, third, and fourth fields, respectively (solid lines indicate lines in odd fields, and broken lines indicate lines in even fields). The sampling is a 4-field sequence and the image compression ratio is approximately 2: 1. Further, in the description of the embodiment of the present invention, a case where the present invention is applied to transmission of a Hi-Vision signal having 1125 scanning lines, a field frequency of 60 Hz, and an interlace ratio of 2: 1 will be described.

The encoder of the present invention will be described first, but only the luminance signal will be described below for simplicity.

FIG. 3 is a schematic block diagram of the entire configuration of the encoder, and FIG. 4 is a block diagram of FIG. 3 showing filters F ₁ (X), F
₂ (Y), F ₃ (Y), F ₄ (Y) and the filter and modulation characteristics of the modulator are shown in FIG.
7 (a) to 7 (i) show XY two-dimensional spectral characteristics.

The luminance component signal of the input video signal is first band-limited by a low-pass filter (LPF) and then subjected to A / D conversion by an A / D converter. The signal spectrum at this time is a spectrum as shown by the fifth illustrated signal (a). This signal is
₁ Filter F ₁ having a horizontal LPF characteristic as shown by (X)
A dynamic system signal having a spectrum characteristic as shown in a fifth illustrated signal (b) through (X). On the other hand, when the difference between the signal (a) and the signal (b) is taken, a horizontal high-frequency component remains and becomes a signal having a spectrum like the fifth illustrated signal (c). Next, the signal (c) is converted into a vertical LPF having characteristics as shown by F ₂ (Y) in FIG.
To a signal having a spectrum like the fifth illustrated signal (d). Further, after passing this signal (d) through a vertical LPF having characteristics as shown by F ₃ (Y) in FIG.
And a signal having a spectrum as shown in signal (e) is obtained.
On the other hand, the difference from the signal (d) is taken to be a signal having a spectrum as shown by the signal (f).

Subsequently, the signal (f) is modulated, and the modulation carrier of the modulator 11 at this time has a horizontal frequency component of zero and a vertical frequency of 1125/2.
TV book carrier. Amplitude-modulating the signal (f) with a modulation carrier whose vertical frequency is the original sampling vertical frequency / 4 can select a carrier phase at which the amplitude of the modulation carrier becomes zero at the line position of one field in the signal. If one field in the signal (for example,
Signals in the odd field are all set to zero, and signals in the other field (for example, the even field) mean signal processing of alternately multiplying (modulating) +1 and -1 for each line and outputting the result. Is not as complicated as left.

After being converted into a signal having a spectrum like the fifth illustrated signal (g) by the modulator 11, the fourth illustrated F ₄ (Y)
Vertical high-pass filter (HPF) with characteristics as shown by
To obtain a signal having a spectrum like the signal (h).
Finally, the signal (h) is added to the signal (e) to obtain a static signal having a spectrum like the signal (i). The static signal (i) is a signal of the dynamic system sent via the upper block F ₁ (X) and two DLY (delay) circuits in the block MIX and the image of the image detected by the separate motion detecting circuit. Mixing is performed by MIX according to the amount of motion, and the block 12 after the block MIX receives the field offset / frame inversion subsampling and sends it to the transmission system 13 to complete the encoding process.

Next, FIG. 6 shows a schematic block diagram of the overall configuration of the decoder according to the present invention, and FIG. 7 shows the block filters F ₅ (X) and F ₆ (Y) used in the configuration shown in FIG. FIG. 8 shows the XY of the signals (j) to (r) of each part in FIG.
Shows two-dimensional spectral characteristics.

In the decoder processing shown in FIG. 6, the processing is different between a still system (still image processing) and a moving system (moving image processing). In the stationary system, first, a frame interpolation is performed by a frame interpolation circuit 21 to remove aliasing components by sub-sampling to obtain a spectrum signal such as a signal (j) shown in FIG.
Next, a horizontal LP having characteristics as shown by F ₅ (X) in the seventh illustration
Via F, a signal having a spectrum like the eighth illustrated signal (k) is obtained. Further, by taking the difference between the signal (k) and the signal (j), a high-frequency component related to the horizontal like the signal (l) is obtained. Further, after passing the signal (l) through the vertical LPF having a characteristic such as F ₆ (Y) shown in FIG. 7, one of the signals (l) outputs a signal having a spectrum such as a signal (m) in addition to the signal (k). And the other obtains a spectral signal having characteristics such as signal (n) obtained by taking the difference from signal (l).

The signal (n) is modulated by the same modulation carrier of the modulator 24 as the modulation carrier 11 of the modulator used on the encoder side. The modulator hardware is therefore very simple, so that the signal (n) is converted into a signal (o) and the filter F
_{6 The} signal is converted into a signal (p) through the vertical LPF having the characteristic of (Y), and is finally added to the signal (m) to complete the static signal processing.

The signal of the dynamic system only needs to be subjected to field interpolation by the field interpolation circuit 22 to remove aliasing components due to sub-sampling. The static system signal and the dynamic system signal are mixed separately according to the amount of motion of the image by the motion detection circuit,
The original analog luminance signal can be restored through D / A conversion by the D / A converter and through the interpolation filter LPF. The above is the processing on the decoder side.

Although the embodiment relating to the encoder and the decoder according to the present invention has been described above, the present invention is not limited to this, and various modifications and changes can be made without departing from the gist defined in the claims. It will be obvious to those skilled in the art.

(Effect of the Invention) In the conventional field offset sub-sampling (MUSE-T system) of the image signal band compression, since the carrier generated by the sub-sampling exists on the plane of the time frequency T = 0 Hz, it is necessary to avoid aliasing distortion. The frequency range of a still image that can be transmitted has been limited to half of the Nyquist frequency range corresponding to a carrier-free original sample generated by subsampling. However, by performing the operation of inverting the sub-sampling pattern between frames as in the present invention, the carrier generated by the sub-sampling on the plane of the time frequency T = 0 Hz is offset in the time frequency direction, and the sub-sampling is further performed. Before performing the modulation (signal processing from signal (f) to (g) in FIG. 5) and the band limiting operation (signal processing from signal (g) to (h) in FIG. 5), the spectrum , The spectral region of a still image that can be transmitted is improved by about 1.5 times, and the time frequency T
= 15Hz motion detection is also possible.

FIG. 9 illustrates the possibility of motion detection on the T = 15 Hz plane. FIG. 9A is a spectrum diagram of the stationary system according to the present invention on the plane of the time frequency T = 0 Hz. When viewed from the plane of the frequency T = 15 Hz, the result is as shown in FIG. 9B. Due to the effect of the modulator 11 in the encoder shown in FIG. 3, the time frequency T = 0 Hz on the plane of the time frequency T = 15 Hz.
The region 31 of the stationary system with a diagonal line at the edge of the surface moves to the region 32, and the region 33 with the diagonal line near X = 0 and Y = 0 has no spectrum of the static system. Is easy to detect the amount of motion.

In addition, it is also possible according to the present invention that the motion amount is not detected using this area, but the motion detection amount detected by another method is converted into a digital signal and inserted into a vertical blanking period of an image and transmitted.

[Brief description of the drawings]

1 shows a sampling pattern according to the present invention, FIG. 2 shows an arrangement diagram of various carriers generated by the sampling pattern, FIG. 3 shows FIGS. 4 and 5, and FIG. FIG. 3 is a schematic block diagram of the overall configuration of the encoder, each block shown in FIG. 3, filters F ₁ (X), F ₂ (Y), F ₃ (Y), F
₄ (Y) and modulator filter and modulation characteristics, third
FIG. 6, FIG. 7, FIG. 8, and FIG. 8 show schematic block diagrams of the entire configuration of the decoder according to the present invention, and FIG. FIG. 9 shows the filter characteristics of the blocks and filters F ₅ (X) and F ₆ (Y) and the XY two-dimensional spectral characteristics of the signals (j) to (r) of the respective parts shown in FIG. 6, and FIG. FIG. 10 is a diagram for explaining a motion detection window according to the present invention on a plane, and FIG. 10 is a diagram showing a sampling pattern according to the MUSE-T system and an arrangement diagram of sampling carriers generated on the plane at a time frequency T = 0 Hz. Show. 11, 24 modulator 12 field offset / frame inversion subsampler 13 transmission system 21 frame interpolation circuit 22 field interpolation circuit

──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Yukihiro Nishida 1-10-11 Kinuta, Setagaya-ku, Tokyo Japan Broadcasting Corporation Research Institute of Broadcasting Technology (58) Field surveyed (Int.Cl. ⁶ , DB name) H04N 7 / 015

Claims (2)

(57) [Claims] An image signal band compression transmitting apparatus for transmitting a high-definition television signal after performing motion adaptive sub-sample band compression on a high-definition television signal, wherein horizontal and vertical high frequency components of a still image transmittable area by original sampling are deleted. Band control means; sampling means for sub-sample band compression by field offset / frame inversion subsampling in which the sampling pattern makes a round in four fields; and prior to the field offset / frame inversion subsampling, the horizontal and vertical high frequency components are A region near the zero frequency of the horizontal and vertical spatial frequencies in the still image two-dimensional spectrum, which is a partial region of the deleted still image two-dimensional spectrum and whose time frequency after the field offset / frame inversion subsampling is the frame frequency / 2. Corresponding to The region is horizontally and vertically deleted in advance by a modulation operation that performs amplitude modulation with a modulation carrier whose horizontal and time frequencies are both zero and a vertical frequency is the original sampling vertical frequency / 4, and a band limiting operation that removes unnecessary sideband components. An image signal band compression transmitting apparatus comprising at least spectrum moving means for moving the image signal to a vertical high frequency side of a high frequency band. 2. A band-compressed image signal receiving apparatus for receiving a high-definition television signal transmitted by an image-signal band-compressing transmitting apparatus according to claim 1, wherein said apparatus performs frame interpolation and performs subsampling. Frame interpolating means for generating a still image two-dimensional spectrum signal from which aliasing components have been removed, and a spectrum shift for moving a part of the still image two-dimensional spectrum, which has been moved in advance at the time of transmission, back to the original area And a means for receiving a compressed band image signal.