JPH08125967A - Image data recording and reproducing device - Google Patents

Abstract

PURPOSE: To make possible variable-speed reproduction with high image quality while simplifying a reproducing circuit by recording input compressed image data after reconstituting them into compressed image data which are composed of small restoration units and satisfy a prescribed image compression format. CONSTITUTION: A slice in the compressed data of an MPEG2 format inputted from a compressed image information input terminal 1 is divided into smaller slices by a silce reconstituting means 2. A recording signal processing means 3 performs recording signal processing to the bit stream and the result is recorded from a recording and reproducing head 4 to a magnetic recording tape 5. At the time of normal-speed reproduction, the signal is reproduced from the magnetic tape 5 by recording and reproducing head 4, processing such as synchronizing detection, error correction and deshuffling is performed by a regenerative signal processing means 6, and the signal of the MPEG2 format is decoded and outputted from a compressed image output terminal 7. This signal is decoded into an image signal by an external MPEG2 decoder and outputted onto a display.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image data recording / reproducing apparatus for recording a digital moving image input in a compressed state.

[0002]

2. Description of the Related Art A highly-compressed image signal (bit stream) used in digital television broadcasting or the like is recorded on a recording medium such as a VTR without decompression, thereby simplifying the VTR device and improving recording efficiency. As an example of such an attempt, there is an image data recording / reproducing apparatus described in Japanese Patent Laid-Open No. 6-86217. In this conventional example, recording / reproduction is performed without expanding the bit stream, and as a result, a signal having the same format as the input signal is output as a reproduction signal, expanded by an external image expansion device, and displayed on a display. .

By the way, in recent years, MPEG (Moving Picture)
MPE being standardized by Experts Group)
Since the G2 system can compress a moving image with high density, it is about to be used for digital television broadcasting. By recording and reproducing the moving image broadcast by the MPEG2 system without expanding it, it is possible to construct a VTR that is inexpensive and has high recording efficiency.

In the MPEG2 system, compression efficiency is improved by using two methods, a method of compressing an image itself and a method of compressing only a difference by referring to other images temporally before and after. A screen compressed using the former is called an intra picture, and a screen compressed using the latter is called a non-intra picture. Non-intra pictures can be compressed into less information. Further, by providing the intra pictures periodically, the reference image can be restored even when the bit stream is interrupted.

Since the VTR records data on the magnetic tape by the helical scan method, only a part of the information recorded on each track can be reproduced by the variable speed reproduction. Therefore, when variable speed reproduction is performed by the VTR in which the bitstream is recorded, a part of the bitstream is divided and reproduced. The MPEG2 system adopts a slice structure so that a reproduced image to some extent can be obtained even when only a part of a bitstream is input (reproduced).

FIG. 5 is an explanatory diagram showing an example of a slice structure. In MPEG2, an image is divided into macroblocks of 16 pixels × 16 pixels, and adjacent macroblocks are collected to form a slice. In this example, an image of 704 × 480 pixels is divided into 30 strip-shaped regions in units of 16 pixels in the vertical direction, and each region is defined as one slice (restoration unit). One slice consists of 44 macroblocks.
Each slice is compressed in macroblock units, and a slice header is added to each slice for transmission. The slice header stores information indicating in which part of the image the slice is arranged, and information indicating a data encoding method. If there is a slice header, it is possible to decode the information in the slice and arrange the pixel information obtained as a result of the decoding on the screen (only the intra slice can be reproduced. In the non-intra slice, the reference image data is reproduced. Images cannot be restored without it).

The slice length can be set to any macroblock length as long as it fits in one horizontal row of the image, but the slice structure of the MPEG2 bit stream used in digital broadcasting is as described above. On the side 1
It is expected that the structure will be one slice per row. This is because if the slice length is made shorter than this, an increase in the code amount due to the slice header cannot be ignored and compression efficiency is deteriorated. In addition, since there is almost no error in the transmission line used for broadcasting, there is no problem even with such a large slice structure.

However, when the VTR is used for variable speed reproduction, the bit stream is divided, so that the reproduced image cannot be obtained unless the data can be restored in units of smaller blocks. For example, when the data is reproduced at 20 times speed, only the data having a length of 1/20 of the track length can be reproduced. If the data of one screen of an intra picture is recorded using three tracks, the data length of one slice is 1/10 of the track length. Therefore, even if the data is reproduced at this speed, it is impossible to reproduce in slice units. Therefore, it cannot be restored at all.

In order to solve such a problem, a technical report of the Institute of Electronics, Information and Communication Engineers [Magnetic Recording] October 21, 1993.
As described on p12 in Japanese MR93-25-29, p7-13, "System study of US ATV compatible digital VTR", the slice is divided into small sync blocks, and additional information indicating the position of the head data is inserted and recorded. It was devised such as to do.

[0010]

According to the above-mentioned conventional example,
If the information can be reproduced in sync block units, the corresponding pixels can be restored, so that it is possible to reproduce a part of the image even at a higher speed.

However, the signal reproduced from the recording medium is
Since only a part of the slice information is included,
In order to be decoded by an external decoder, it is necessary to perform processing such as interpolating a portion that has not been reproduced. Therefore, it is necessary to once decode the signal reproduced from the recording medium, perform processing such as interpolation, and further encode and output the signal by a predetermined method.

An object of the present invention is to provide an image data recording / reproducing apparatus capable of high-speed variable speed reproduction even if a bit stream is recorded on a recording medium without being expanded.

[0013]

In the MPEG2 system, the lower limit of slice length is one macroblock. In other words, even if the bit stream transmitted by digital broadcasting has a slice structure composed of many macroblocks, each slice is divided into subslices consisting of a small number of macroblocks, and a slice header is added to each subslice. The bit stream satisfying the MPEG2 method is reconstructed by performing the processing such as the above.

That is, the image data recording / reproducing apparatus according to the present invention receives the compressed image data compressed by a predetermined image compression format, and records the compressed image data on a predetermined recording medium. Restoration unit reconstructing means for reconstructing the restoration unit block of image data into compressed image data composed of a smaller restoration unit block group and satisfying the predetermined image compression format, and recording it on a predetermined recording medium. Be prepared.

The predetermined image compression format is M
The PEG1 or MPEG2 system is used.

Further, in the restoration unit reconstructing means, the predetermined image compression format is MPEG1 or MPEG2 system, and the restoration unit block before reconstruction is MPEG1, M.
It is an intra slice defined by the PEG2 method.

In the restoration unit reconstructing means, the predetermined image compression format is MPEG1 or MPEG2, and the restoration unit block before reconstruction is MPEG1, MP.
When a non-intra slice defined by the EG2 method includes an intra macroblock, it is reconstructed into an intra slice composed of intra macroblocks and a non-intra slice including non-intra macroblocks.

The above-mentioned MPEG1 is standardly selected by ISO / IEC as (ISO / IEC11172), and MPEG2 is also specified by ISO / IEC (ISO
The standard is being selected as SO / IEC 13818). ISO: the International Organization f
orstandardization, IEC: the International Elect
It is a rotechnical Commission.

[0019]

In the present invention, since the bit stream has a shorter slice length, even if the bit stream is recorded on the magnetic tape without being expanded, a higher speed variable speed reproduction can be realized. Becomes Further, since it is not necessary to perform processing for removing additional information from reproduced data or to reconstruct slices by interpolating missing data, reproduction processing can be performed more easily. Further, in the MPEG2 system, since the bit stream input to the decoder does not have to cover the entire screen, the external decoder can appropriately perform the processing even if the VTR does not interpolate and output the missing slice due to the variable speed reproduction. To restore and output to the display. Further, since the bit stream output in the normal speed reproduction also complies with MPEG2, it is not necessary to reduce extra header information during reproduction. In this way, it is possible to easily output a signal conforming to the MPEG2 system in both variable speed reproduction and normal reproduction.

[0020]

1 is a block diagram showing the configuration of an embodiment in which the present invention is applied to a digital VTR.

The configuration of FIG. 1 will be described together with its operation. The slice in the compressed image data of MPEG2 format input from the compressed image information input terminal 1 is divided into smaller slices by the slice reconstructing unit 2 as the reconstructing unit reconstructing unit. The size after division is the method of making each slice after division include the same number of macroblocks,
Any division method may be used, including a method of making the number of bits of each slice after division almost equal. The bit stream obtained as a result of the division is also processed so as to comply with the MPEG2 format. For this bitstream,
The recording signal processing means 3 performs predetermined recording signal processing such as shuffling, addition of error correction code, addition of synchronization signal, recording modulation, etc., and recording from the recording / reproducing head 4 to the magnetic tape 5 as a recording medium.

During normal speed reproduction, the recording / reproducing head 4 is used to reproduce a signal from the magnetic tape 5, and the reproduction signal processing means 6 carries out processing such as synchronization detection, error correction, and deshuffling to obtain an MPEG2 format signal. Is restored and output from the compressed image information output terminal 7. This signal is restored to an image signal by an MPEG2 decoder prepared externally and output to a display.

The operation of the slice reconstructing means 2 in FIG.
This will be described in more detail with reference to FIGS. 2, 3 and 4.

FIG. 2 is a block diagram showing an example of the configuration of the slice reconstructing means 2, and FIG. 3 is a diagram showing a bitstream before and after processing performed by the slice reconstructing means 2. FIG. 4 is a diagram showing an example of a slice structure in the MPEG2 format.

In FIG. 4, slices are roughly classified into slice headers and slice data. The slice header consists of various parameters, and the slice data consists of macroblocks.

The meanings of the parameters included in the slice header are as follows. The slice start code is a 32-bit data string that indicates the beginning of the slice.
Hexadecimal number is 000001 ^** . The last two characters (= 8bit) are
Indicates where this slice is arranged in the vertical direction on the screen. Since this data string is a code that does not occur in other data, decoding can be restarted by searching the slice start code even if the bit stream is interrupted. The Q scale code is a parameter for quantizing data. The value of the Q scale code is first defined in the slice header, but it can be changed in macroblock units. When the Q scale code is changed in a macro block on the way, the value of the Q scale code is described in the macro block. The intra slice flag indicates whether all the data in this slice are intra macroblocks. At the time of variable speed reproduction, the intra slice can be extracted based on this intra slice flag. The slice header may include various parameters in addition to these, but the description is omitted because it does not directly relate to the content of the present invention.

The meaning of the parameters included in the macroblock is as follows. The macroblock address increment is a parameter that indicates where the macroblock is arranged in the horizontal direction on the screen. The macroblock at the beginning of the slice indicates the number of macroblocks from the left end of the screen, and the other macroblocks indicate the previous macroblock. Indicates the distance from the block. The value is 1 when the macroblocks are continuously arranged, but is a value greater than 1 when the macroblocks are encoded by skipping (because skip is not allowed in intra slices, the value is 1). In other words, the macroblock at the head of the slice indicates the absolute position in the horizontal direction, and the other macroblocks indicate the relative position to the previous macroblock.

The macroblock type includes information indicating intra / non-intra, information indicating whether or not the Q scale code is changed, and the like. When there is a change in the Q scale, the Q scale code is also added. In the macroblock data, quantized DCT information is stored in the case of an intra macroblock, and quantized DCT information and motion vectors are stored in the non-intra macroblock. Filling information is added at the end of the slice so that the total number of bits of the slice thus configured is a multiple of 8.

In order to divide the slice having such a structure into a plurality of slices, the following processing is required.
First, the slice header is read. Next, points to be divided in the slice data are detected. The slice head can be searched by the slice header, but since the inside of the slice is variable-length coded, the variable-length code is decoded and the boundary between the macroblocks to be divided is searched for as the division point. Next, a slice header is added to the beginning of each divided slice.

Explaining FIG. 3 as an example, slice data 0, which is a restoration unit block, is divided into two smaller restoration unit blocks, and a slice header copy (0-
b) is inserted. However, Q in slice header 0-b
The scale code is the value of the Q scale code of the leading macroblock of the subsequent slice data. Since the vertical coordinate in the screen of the divided slice does not change, it is not necessary to change the slice start code. Further, if the original slice is a slice including a non-intra macroblock and the slice after division is composed of only intra macroblocks, the intra slice flag in the slice header also needs to be modified.

Next, as a result of the division, the macro block address increment of the macro block which becomes the head of the slice is converted from the relative address to the absolute address. Next, if the total bit length of the slice obtained as a result of the division is not a multiple of 8, the filling information is added or deleted to make the total bit length a multiple of 8.

FIG. 2 is a block diagram showing details of an example of the slice reconstructing means 2 for performing the above processing. The configuration and operation will be described below.

From the information input from the compressed image information input terminal 1, the slice header detection circuit 101 searches for a slice start code, decodes the variable length coded data that follows, and slice header (Q scale code, intra code). Slice flags). Further, the subsequent bitstream is processed by the macroblock head detection circuit 104
Decoding is performed to detect the head of the macroblock from the slice data. The number of macroblocks in the slice after reconstruction is counted by the division point determination circuit 105, and the point at which the predetermined number of macroblocks is reached is taken as the division point. Further, the Q scale code of the head macroblock of the slice after the division is calculated, and the Q scale code correction circuit 102 corrects the Q scale code of the slice header to be inserted. If the slice after reconstruction does not include a non-intra macroblock, the intra slice correction circuit 10
The intra slice flag of the slice header inserted in 3 is set to a predetermined value. The slice header corrected in this way is subjected to variable length coding, and then the slice header insertion circuit 106 inserts it at the division points.

Next, in order to correct the macroblock address increment flag of the slice top macroblock after reconstruction, the macroblock address increment detection circuit 107 decodes the macroblock address increment information. Since this information is a relative address except for the slice head macroblock, the macroblock address counter 108 converts it to an absolute address. Then, increment the macroblock address of the new macroblock that becomes the slice head,
Macro block address increment correction circuit 109
Convert to an absolute address with.

Filling information is added or deleted by the byte arrangement adjusting circuit 110 so that the total number of bits of the slice after reconstruction obtained as a result of the above processing is a multiple of 8. With the above processing, it is possible to create a bitstream that complies with the MPEG2 format and has a slice structure that is shorter than the input bitstream.

In the present embodiment, an example in which it is realized by using hardware has been described, but it is also possible to perform the same processing by software processing.

[0037]

As described above in detail, the present invention is applied to the input compressed image in the image data recording / reproducing apparatus for inputting the compressed image data compressed by the predetermined image compression format and recording it on the predetermined recording medium. A restoring unit reconstructing unit configured to reconfigure the restoring unit block of data into compressed image data which is composed of a smaller restoring unit block group and which satisfies the predetermined image compression format, and records it on a predetermined recording medium. Therefore, since it is divided into smaller restoration units so as to satisfy the image compression format, it is not necessary to convert the format of the reproduced signal, and the reproducing circuit becomes simple.

Further, according to the present invention, since a bitstream conforming to the MPEG2 format and having a restoration unit structure shorter than the restoration unit structure of the input bitstream is created and recorded on the recording medium, variable speed reproduction is performed. If this is done, it becomes possible to regenerate more restoration units. As a result, variable speed reproduction with higher image quality can be realized.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment in which the present invention is applied to a digital VTR.

FIG. 2 is a block diagram showing details of the configuration of a slice reconstruction circuit in the embodiment of FIG.

FIG. 3 is a diagram showing a bitstream before and after processing performed by a slice reconstructing unit according to the present invention.

FIG. 4 is a diagram showing an example of a slice structure in the MPEG2 format according to the present invention.

FIG. 5 is an explanatory diagram showing an example of a conventional slice structure.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 compressed image information input terminal 2 slice reconstructing means 3 recording signal processing means 4 recording / reproducing tape 5 recording medium 6 reproducing signal processing means 7 compressed image information output terminal

 ─────────────────────────────────────────────────── ─── Continued Front Page (72) Inventor Toshifumi Sakaguchi 4-36-19 Yoyogi, Shibuya-ku, Tokyo Within Graphics Communication Laboratories, Inc. (72) Inventor Yusumi Wataya 4-chome Yoyogi, Shibuya-ku, Tokyo 36th-19th Stock Company, Graphics Communication Laboratories

Claims (4)

[Claims] 1. An image data recording / reproducing apparatus which receives compressed image data compressed by a predetermined image compression format and records the compressed image data on a predetermined recording medium. An image data recording / reproducing device comprising a restoration unit reconstructing unit which is composed of a unit block group and which is reconstructed into compressed image data satisfying the predetermined image compression format and recorded on a predetermined recording medium. apparatus. 2. The predetermined image compression format is MPEG
2. The image data recording / reproducing apparatus according to claim 1, wherein the image data recording / reproducing apparatus is of the 1 or MPEG2 system. 3. The restoration unit reconstructing means uses a predetermined image compression format of MPEG1 or MPEG2 system, and the reconstruction unit block before reconstruction is MPEG1, MPE.
The image data recording / reproducing apparatus according to claim 1, wherein the image data recording / reproducing apparatus is an intra slice defined by the G2 system. 4. The restoration unit reconstructing means uses a predetermined image compression format of MPEG1 or MPEG2 system, and the reconstruction unit block before reconstruction is MPEG1, MPE.
The non-intra slice defined by the G2 method, when including an intra macroblock, is reconfigured into an intra slice composed of the intra macroblock and a non-intra slice including the non-intra macroblock. Image data recording / reproducing apparatus.