JP3407943B2 - Video signal recording device - Google Patents

Description

Detailed Description of the Invention

[0001]

The present invention relates to an HDTV signal (High D
The present invention relates to an encoding device that processes a plurality of pixels in blocks when high-efficiency encoding (efinition Television signal), particularly sharing with an SDTV signal (abbreviation of Standard Television signal such as NTSC signal or PAL signal). The present invention relates to a video signal recording device for performing digital recording for the purpose.

[0002]

2. Description of the Related Art An apparatus for encoding and recording and reproducing a video signal with high efficiency, such as a digital VTR, is constructed as shown in FIG. 16, and its construction will be briefly described with reference to FIG. Generally, a digital VTR rearranges blocks of a certain size as a unit (in the block shuffling 101) and then performs high efficiency encoding (in the high efficiency encoding circuit 102) in order to reduce the amount of information recorded on the tape. ). Next, a code (parity) for error correction is added to the high efficiency coded data (in the error correction coding circuit 103). Next, after adding a synchronizing signal / ID signal for recording / reproduction (Sync, in the ID adding circuit 104), it is modulated so as to suppress the DC component (in the modulating circuit 105) and recorded on the tape (recording medium 106). .

On the reproducing side, error correction is performed based on the demodulated data (in the demodulation circuit 107), and then the encoded block is decoded as a unit (error correction decoding 109).
At). At this time, the portion that could not be corrected by the correction circuit is not decoded but is corrected by the decoding / correction circuit 110. After that, the blocks are rearranged in the reverse order of the recording side (in the block deshuffling circuit 111) to reproduce the original video signal. The above is the basic configuration of the digital VTR using high efficiency coding.

A method of recording and reproducing an SDTV signal and an HDTV signal by the same mechanism system by using such a device is being studied. As one of the methods, since the information amount of the HDTV signal is much larger than that of the SDTV signal, the code amount after high efficiency encoding is set to twice that of the SDTV signal, and the encoded signal is transmitted at the tape feeding speed of SDTV. There is an idea that recording is performed with almost twice the recording rate on the magnetic tape by recording at twice the signal recording.

[0005]

DISCLOSURE OF THE INVENTION Problems to be Solved by the Invention
The V method has different specifications in Japan, the United States, and Europe. In Japan, the 1125/60 method and in the United States, 105 method.
The 0/60 method and the 1250/50 method in Europe are influential. (Note that in the United States and Europe, the final decision has not yet been made.) For example, in the Japanese method defined by BTA-S002, since there are 1035 effective lines / frame and 1920 effective pixels / line, 8 pixels in the horizontal direction. × When orthogonal transformation is performed with 8 lines in the vertical direction as one block,
There is a problem that fractions occur in the vertical direction.

In addition, the code amount after high-efficiency coding is SDTV.
In the process of doubling the signal, it is desirable that the number of lines orthogonally transformed is twice that of the SDTV signal.
When the field frequency is 60 Hz (including 59.94 Hz), there are 960 lines of HDTV signal for 480 lines of SDTV signal, and when the field frequency is 50 Hz, there are 1152 lines of HDTV signal for 576 lines of SDTV signal. Is good. In that respect, 1050 /
60 system, 1250/50 system are NTSC:
Since it is composed of twice the number of scanning lines of the 525/60 system and the PAL: 625/50 system, it is easy to set the number of lines to be orthogonally transformed to the above-mentioned value, while the number of effective lines is already 1035. Specified as 1125/6
In the 0 system, if the orthogonal transform processing is performed as it is, there is a problem that the processing process at the time of high efficiency coding becomes complicated.

[0007]

In the present invention SUMMARY OF] is 1 off
The HDTV video signal of Laem is used for the effective pixel number and effective line.
And a means for converting to a video signal whose number is an integer multiple of 8.
The converted 1-frame signal is output horizontally and vertically
Means for using 8 pixels x 8 lines as one DCT block
And a plurality of corresponding DCT blocks on the screen.
Means for constructing the black block and L (L is an integer)
High-efficiency coding is performed on the black block to fix the fixed length.
Generate P recording blocks (P is an integer such that P ≦ L)
Means for recording the integer number of recording blocks on a recording medium
And a video signal recording apparatus comprising:
The number of macroblocks and recording blocks per frame
Are K when recording NTSC or PAL signals, respectively.
Times, KP / L times (K is an integer), and the recording block
The recording medium is recorded on the recording medium,
Record at KP / L times when recording C or PAL signals.
Therefore, when recording HDTV video signals, SDTV signal recording
It is possible to share a lot of processing with recording.

[0008]

In the above structure, when recording the HDTV signal,
The number of macroblocks per frame by SDT
Pixels of video signal, for example, four times as large as when recording V signal
The number is changed by processing such as thinning, and then the macro block
When creating a recording block from the
When one recording block is generated from the lock, one frame
The number of recording blocks per frame is the same as when recording SDTV signals.
Doubled. When recording this signal on the recording medium,
Doubled, that is, if the recording medium is tape,
Error correction processing by recording the feed rate of the loop twice
Can be shared with HDTV and SDTV, and for tape
The recording rates are the same, and the processing is easy.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIGS. FIG. 1 is an example of a video signal encoding apparatus according to the present invention. In the figure, 1
Is thinning means, 2 is blocking means, 3 is orthogonal transformation means, and 4 is code amount control means. The operation of the video signal encoding apparatus of the embodiment of the present invention configured as described above will be described below. The thinning means 1 thins the effective pixels and the effective lines from the input digital video signal according to the input signal (details will be described later). Next, the thinned signal is divided into 8 pixels by the blocking means 2.
Block 8 lines. The orthogonal transform means 3 uses a well-known DCT (Discrete Cosi).
gn Transform) and the like to perform orthogonal transformation in units of the blocks. The code amount control means 4 quantizes the orthogonally transformed coefficient data according to the degree of importance for representing an image, and further performs variable length coding such as Huffman coding to reduce the amount of information, and The generated code amount is controlled to be constant for each one or a plurality of blocks to form a recording block, and the total code amount generated for the entire frame is equal to the code amount after encoding the SDTV signal, which is 2
Control to double. In this way, the recording block configured is subjected to recording signal processing such as error correction coding and modulation (not shown), and recorded on the magnetic tape.

Below, the input HDTV signal is 1125/60.
This will be specifically described by taking the case of the system as an example. Figure 1
It is a schematic diagram showing the input signal of a 125/60 system. As shown in the figure, the effective area of one frame is a luminance signal (Y) 1920 pixels × 1035 lines, and a color difference signal (U,
V) Consists of 960 pixels × 1035 lines.

The thinning-out means 1 shown in FIG. 1 thins out pixels and lines using a well-known digital filter, and further performs so-called line-sequentialization for color difference signals, as shown in FIG. Get the signal. As shown in FIG. 3, the effective area of the signals thinned out in this way is composed of a luminance signal of 1440 pixels × 960 lines and a color difference signal of 720 pixels × 480 lines.

The effective area is a 4: 1: 1 component signal effective area of the NTSC signal shown in FIG. 4, luminance signal 720 pixels × 480 lines, color difference signal 180 pixels ×
The number of pixels is exactly four times that of 480 lines.

Since the object of the present invention is to thin out the effective pixels or effective lines in this way and to make them suitable for the processing of the blocking means 2 and below, the characteristics of the digital filter and the method of line-sequencing. And the configurations and the order thereof are not particularly limited. Further, as a method of thinning lines, it is possible to appropriately combine the effective lines at the top and bottom of the screen and delete a total of 75 lines, for example, 35 lines at the upper side and 40 lines at the lower side, to obtain 960 lines. This is not preferable because the original information is significantly lost.

The signal thinned out in this way is converted into a DCT block of 8 pixels × 8 lines by the blocking means 2. The number of DCT blocks generated in this way is also NTSC
It is four times the number of DCT blocks of the signal (see FIG. 5).

The orthogonal transform means 3 performs orthogonal transform by DCT in units of the DCT block. The code amount control means 4 adaptively quantizes the orthogonally-transformed coefficient data by switching the quantization table according to the degree of importance for restoring an image, and further variable length such as Huffman coding. Encoding is performed to reduce the amount of information. At this time, the generated code amount is controlled to be constant for each one or a plurality of DCT blocks. However, in the component signal as shown in FIG. Since both of them are necessary, the DCT block of the corresponding luminance signal and chrominance signal is regarded as one unit on the screen (hereinafter referred to as macroblock), and the code amount generated for each one or a plurality of macroblocks is constant. In addition to the above, the total code amount generated in the entire frame is controlled to be twice the code amount of the NTSC signal after encoding.

In the structure of the DCT block shown in FIG. 5, a total of 6 DCT blocks, each of which has four luminance signals and one color difference signal, covers a pixel corresponding to an area on the screen. The macroblock shown in FIG. 6 is constructed from the individual DCT blocks. From the figure, the number of macroblocks in one frame is 90 × 60, that is, 540.
The number is 0, which is four times as large as the NTSC signal composed of 1350 macroblocks.

Now, a method of constructing a recording block from the code whose information amount has been reduced will be described. Here, in the code amount control of the NTSC signal, a case where the data of one macroblock is stored in one recording block will be additionally described. Since this is a problem of the ratio between the number of macroblocks and the number of recording blocks, the same applies even if the data of five macroblocks are contained in five recording blocks.

When the recording blocks are formed in the NTSC signal at such a ratio of 1: 1 as described above, the number of recording blocks per frame is 1350.

As a method of recording and reproducing the NTSC signal and the HDTV signal by the same mechanism system, the tape feeding speed is set to NTS.
There is a method in which recording is performed at twice the C signal recording, so that the recording rate on the magnetic tape is hardly changed. Therefore, without changing the size of the recording block,
If the number of recording blocks is doubled, the processes after the error correction coding can be easily performed. As described above, since the HDTV signal is composed of 5400 macroblocks,
As shown in FIG. 7, if one recording block is composed of data of two macro blocks, the number of recording blocks per frame becomes 2700, which can satisfy the above condition.

Further, since the above method is a high-efficiency coding method utilizing only intra-frame correlation, the NTS
Since the compression rate is simply doubled as compared with the C signal, the image quality may be insufficient. In that case, in addition to intra-frame correlation, it is necessary to perform coding using inter-frame correlation. For example, when performing coding in units of two frames, as described in Japanese Patent Application No. 5-191289 filed by the applicant, the intra-frame correlation of macroblocks at the same position on the screen is used. There is a method of storing the encoded data and the encoded data using the inter-frame correlation in the same recording block. At this time, as shown in FIG. 8, the encoded data of the macroblock to be paired is 1
When assigned to one recording block, the total number of recording blocks in two frames is 5,400, that is, the number of recording blocks in one frame is 2,700.
The above conditions can be satisfied.

As described above, the size of the recording block is changed for the NTSC signal regardless of the high-efficiency coding means by thinning out the effective pixels and the effective lines by the appropriate thinning means before the block formation. Instead, since the number of recording blocks per frame can be doubled, recording processing such as error correction coding and modulation can be easily realized by parallel processing of NTSC signal processing.

The above is the outline of the present invention. Next, various modifications of the present invention will be described. The second embodiment relates to the number of samples in the horizontal direction. In the first embodiment, the case where the luminance signal is thinned out to 1440 pixels × 960 lines and the color difference signal is 720 pixels × 480 lines has been described. However, in order to reduce the influence of block distortion and the like due to the coding using the DCT and the circuit. There is a method of increasing the thinning amount of pixels in consideration of the scale. As shown in FIG. 9, the luminance signal is 1080 pixels × 960 lines, and the color difference signal 3 is
A case of thinning out to 60 pixels × 480 lines will be described.

As described above, when the DCT blocks are formed from the thinned signals, the number of DCT blocks per frame is 8/3 times that of the NTSC signal (FIG. 10).
From now on, as in the first embodiment, considering the configuration of the macroblock, 6 luminance signals and 1 color difference signal are provided.
Since a total of 8 DCT blocks cover pixels corresponding to an area on the screen, the 8 DCT blocks
From the T block, the configuration shown in FIG. 11 is obtained. From the figure,
The number of macroblocks in one frame is 45 × 60, that is, 2700, which is twice the number of NTSC signals.

Therefore, when high-efficiency encoding is performed using only intra-frame correlation, if the data of one macroblock is stored in one recording block, the number of recording blocks per frame will be 2700,
It can be twice the NTSC signal. On the other hand, in the case of encoding using intra-frame correlation and inter-frame correlation, as shown in FIG. 12, when the coded data of the macroblocks forming a pair are assigned to two recording blocks, two frames are recorded. The total number of recording blocks is 5,400.
That is, the number of recording blocks per one frame is 2,700, and the above condition can be satisfied.
As described above, the present invention can be applied even if the number of samples in the horizontal direction is appropriately set in consideration of the macroblock configuration.

The third embodiment relates to various input signals. As described in the related art, the HDTV system includes 1050 /, in addition to the 1125/60 system (Japan).
60 system (US), 1250/60 system (Europe), etc. In the 1050/60 method, first, the effective line is set to 9
An appropriate vertical blanking period is determined so as to have 60 lines, and the color difference signals are line-sequentially set to 480 lines. On the other hand, the effective pixels are thinned out by the digital filter as described above, and the luminance signal is 1440 pixel lines × 960 lines and the color difference signal is 720 pixels × 480 as shown in FIG.
Lines, or luminance signals of 1080 pixels × 960 lines and color difference signals of 360 pixels × 480 lines, may be encoded as in the first and second embodiments, respectively.

Further, in the 1250/60 system, similarly,
An appropriate vertical blanking period is determined so that the number of effective lines is 1152, and the color difference signals are line-sequentially set to 576 lines. On the other hand, the effective pixels are thinned out by the digital filter as described above, and as shown in FIG.
40 pixel lines x 1152 lines, color difference signal 720 pixels x 576 lines, or luminance signal 1080 pixels x 1
Assuming that there are 152 lines and 360 pixels of color difference signals × 576 lines, with respect to the effective area of the 4: 2: 0 component signal of the PAL signal shown in FIG. 15, luminance signals of 720 pixels × 576 lines, and color difference signals of 360 pixels × 288 lines. 1125/60 system and NTS described in the first and second embodiments.
Since the same relationship as the relationship of the C signal is established, the same encoding as in the first and second embodiments may be performed.

As described above, the field frequency is 60 Hz (59.94H) even for various input HDTV signals.
(including z), the effective line is 960 lines, 50H
In the case of z, the present invention is applicable if the number of samples in the horizontal direction is appropriately set in consideration of the macroblock configuration so that the number of samples in the horizontal direction is thinned out (interpolated in some cases) so as to be 1152 lines. it can.

[0028]

As described above , according to the present invention.
For example, if the number of macroblocks per frame is SDTV
The number of pixels of the video signal is thinned out so that it is K times the signal recording
Change from the macro block,
L macroblocks when generating recording blocks
When P recording blocks are generated from
The number of recording blocks of KP / L at SDTV signal recording
Doubled. When recording this signal on the recording medium,
To KP / L times, that is, whether the recording medium is tape.
For example, by recording the tape feed speed at KP / L times,
Error correction processing can be shared by HDTV and SDTV, and
The same recording rate for each group makes processing easier
It

[0029]

[0030]

[0031]

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of a video signal encoding apparatus of the present invention.

FIG. 2 is a schematic diagram showing an HDTV signal of 1125/60 system (Japan).

FIG. 3 is a schematic diagram showing an HDTV signal after thinning-out according to the first embodiment.

FIG. 4 is a schematic diagram showing an effective area of a 4: 1: 1 component signal of an NTSC signal.

5 is a schematic diagram showing a DCT block generated from the signal of FIG.

6 is a schematic diagram showing a macroblock composed of the DCT block of FIG.

FIG. 7 is a configuration diagram of an example of a recording block in which encoded data using only intra-frame correlation is stored according to the first embodiment.

FIG. 8 is a configuration diagram of an example of a recording block in which coded data using intra-frame correlation and inter-frame correlation according to the first embodiment is stored.

FIG. 9 is a schematic diagram showing an HDTV signal after thinning-out according to the second embodiment.

10 is a schematic diagram showing a DCT block generated from the signals of FIG.

11 is a schematic diagram showing a macro block composed of the DCT blocks shown in FIG.

FIG. 12 is a configuration diagram of an example of a recording block in which encoded data according to the second embodiment is stored.

FIG. 13 is a schematic diagram showing a 1050/60 HDTV signal after being thinned out according to the third embodiment.

FIG. 14 is a schematic diagram showing a signal after thinning out a 1250/50 type HDTV signal according to a third embodiment.

FIG. 15 is a schematic diagram showing an effective area of a 4: 2: 0 component signal of a PAL signal.

FIG. 16 is a block diagram of a digital VTR using high efficiency coding.

[Explanation of symbols]

1 thinning means 2 Blocking means 3 Orthogonal transformation means 4 Code amount control means

─────────────────────────────────────────────────── ─── Continuation of front page (72) Inventor Masaru Yoshida 22-22 Nagaikecho, Abeno-ku, Osaka-shi, Osaka Within Sharp Corporation (72) Inventor Akiyoshi Nagao 22-22 Nagaike-cho, Abeno-ku, Osaka, Osaka Sharp shares In-company (72) Inventor Futoshi Shimizu 22-22 Nagaike-cho, Abeno-ku, Osaka-shi, Osaka (56) References JP-A-4-86182 (JP, A) JP-A 5-183935 (JP, A) ) JP-A-5-234260 (JP, A) JP-A-6-38239 (JP, A) JP-A-6-334967 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) H04N 5/76-5/956

Claims (2)

(57) [Claims] 1. An effective 1-frame HDTV video signal
For video signals in which the number of pixels and the number of effective lines are integer multiples of 8
Means for converting, and the converted one-frame signal for each of horizontal and vertical
Means for making 8 pixels x 8 lines into one DCT block
And a macro from the plurality of corresponding DCT blocks on the screen.
A means for constructing a block and a high efficiency code for L (L is an integer) macroblocks
Of fixed-length P (P is an integer such that P ≦ L)
Means for generating recording blocks, and means for recording the integer number of recording blocks on a recording medium
A video signal recording device comprising: a macro block for one frame; and a recording block.
The number of marks is the same when recording NTSC or PAL signals.
And KP / L times (K is an integer)
The recording medium is recorded on the recording medium, the feeding speed of the recording medium is N
Record at KP / L times when recording TSC or PAL signals
A video signal recording device characterized by the following. 2. A means for generating the recording block
In addition to intra-frame correlation, inter-frame correlation was used.
The video according to claim 1, wherein the video is encoded.
Image signal recording device.