JPH08154230A - Method for storing moving image coded data on medium - Google Patents

Abstract

PURPOSE: To realize special reproduction such as in fast forwarding or fast rewinding without needing intermediate processing by a CPU by having only to read sequentially data from a storage medium when moving image coding data for special reproduction are stored on a medium in the case of reproduction of image data. CONSTITUTION: Coded data of a moving image are stored on a storage medium 1, and a reproduction means 2 accesses the coding data directly or via a communication line 3 to reproduce a moving image in this system. As the medium storage method of moving image coding data, coding data for usual reproduction and coding data for special reproduction are individually provided on the storage medium and either of data is accessed on request to selectively execute the usual reproduction and the special reproduction optionally.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of storing moving picture coded data on a medium, and more particularly to a method of storing moving picture coded data for special reproduction on a medium.

The demand for image data has increased remarkably, and therefore it is required to record the image data in a large-capacity storage medium so that the image data can be delivered via a communication network on demand. .

Since image data has a large amount of information, when it is recorded, it is often compressed and recorded by high efficiency coding. In this case, the data storage method is not limited to normal reproduction but It is desirable that special playback such as fast-forward and fast-reverse is also possible.

[0004]

2. Description of the Related Art Since image data has a huge amount of information,
When recording on a storage medium such as a hard disk, a magneto-optical disk (MO), a compact disk ROM (CD-ROM), etc., it is often the case that high efficiency encoding is performed and then compression is performed (however, without compression, Sometimes recorded as is). Recently, in this regard, ITU-T
(International Telecommunication Union-Telecommun
ication Standardization Sector). 261 and I
SO (International Organization for Standardizati
International standardization methods such as on) MPEG are also being established.

The recorded and encoded image data is
A moving image is reproduced by being accessed directly from a storage medium or via a communication line. Recently, it has been expected to develop a system such as VOD (Video on Demand) that can access any program at any time at home.

The coded data of an image recorded by a normal storage method can be reproduced normally if read out sequentially. However, not only normal playback,
There is a demand for data that can be trick-played, such as fast-forward and fast-reverse, but no suitable storage method has hitherto been known.

[0007]

SUMMARY OF THE INVENTION In order to perform high speed reproduction from encoded data of an image by a normal storage method,
It is necessary to randomly access the data at random intervals, which increases the access time such as seek time, so that an ultra-high-speed reproducible storage medium is required. Therefore, although it was possible in principle, it was not possible in practice.

Further, in order to read out only the necessary data from the stored data, the data is once read out and then only the necessary data is extracted, or the position where the necessary data is recorded on the storage medium. It is necessary to separately prepare a file that indicates this and read data while referencing this file, but this requires processing by the central processing unit (CPU), etc., and high-speed access is impossible. There was a problem that was.

The present invention is intended to solve the above-mentioned problems of the prior art, and only by reading out the data from the storage medium almost sequentially, fast-forwarding with almost no intermediate processing by the CPU or the like is required. It is an object of the present invention to provide a medium storage method for encoded moving image data, which enables special reproduction such as reproduction, fast-reverse reproduction, and high-resolution still image reproduction.

[0010]

FIG. 1 conceptually shows a system configuration of the present invention, in which 1 is a storage medium for storing encoded data, and 2 is reproducing means for reproducing encoded data. 3 is a communication line connecting the storage medium 1 and the reproducing means 2.

(1) In a system for reproducing moving images by accumulating coded data of a moving image in the storage medium 1 and accessing the coded data from the reproducing means 2 directly or via the communication line 3, the storage medium The normal reproduction coded data and the special reproduction coded data are separately provided above, and either the normal reproduction or the special reproduction is arbitrarily selected by accessing either data according to the request. And make it feasible.

(2) In the case of (1), the coded data for special reproduction is coded data for fast-forward reproduction, coded data for fast-reverse reproduction, or coded data for high-resolution still image reproduction. , Or encoded data composed of any combination thereof.

(3) In the case of (2), a plurality of types of coded data for double speed reproduction are provided as coded data for fast forward reproduction or fast reverse reproduction, and the coded data for double speed reproduction required in accordance with the request is provided. A desired double speed reproduction is performed by accessing only the encoded data.

(4) In the case of (2), the coded data for fast-forward reproduction or fast-reverse reproduction has coded data at a higher compression rate than that of normal reproduction.

(5) In the case of (2), the coded data of the screen which is intra-frame coded every n (n ≧ 1) screens is used as the coded data for the fast forward reproduction or the fast reverse reproduction.

(6) In a system for reproducing moving images by accumulating coded data of a moving image in the storage medium 1 and accessing the coded data directly from the reproducing means 2 or through the communication line 3, the storage medium In the above, the same image sequence (program) has different k (k ≧ k) in which one of the average bit rate, the resolution, and the frame rate is different.
2) By accumulating types of encoded data, a moving image having a desired average bit rate or resolution or frame rate is selected and can be reproduced.

[0017]

In FIG. 1, (a) shows a case where the storage medium is directly accessed, and (b) shows a case where the storage medium is accessed through a communication line.

The system shown in FIG. 1 (a) is directly accessed from a storage medium such as a hard disk of a personal computer in response to a request.
The case of reading data corresponds to this. The system shown in (b) directly accesses a storage medium via a communication line to reproduce it, and a VOD system or the like corresponds thereto.

FIG. 2 shows a structure of image data on a storage medium, and shows a method of storing moving image data in consideration of normal reproduction and special reproduction. (1) is normal playback data, (2) is 5x forward playback data, and (3) is 20
Double speed forward reproduction data, (4) is 5 times reverse reproduction data, (5) is 20 times reverse reproduction data, and (6) is high resolution still image data.

(1) In the present invention, moving image data for normal reproduction and moving image data for special reproduction are separately stored on a storage medium, and when normal reproduction is requested, normal reproduction is performed. When a special reproduction request is made for moving image data for playback, it is possible to sequentially select the normal reproduction and the special reproduction by sequentially reading out the moving image data for special reproduction. Become. However, the amount of accumulated data increases by the amount of data for special reproduction.

(2) In this case, as coded data for special reproduction, coded data for fast-forward reproduction, coded data for fast-reverse reproduction, coded data for high-resolution still image reproduction, or these It is possible to have encoded data consisting of any combination of

(3) In this case, by having a plurality of types of coded data for double speed reproduction as coded data for fast forward reproduction or fast reverse reproduction, the coded data for double speed reproduction required in accordance with the request can be obtained. By accessing only the data, the desired double speed reproduction can be performed.

(4) In this case, the encoded data for the fast-forward reproduction or the fast-reverse reproduction may have the encoded data encoded at a higher compression rate than the normal reproduction, and the accumulated data is stored. Data can be reduced.

(5) In this case, as the coded data for the fast-forward reproduction or the fast-reverse reproduction, it is possible to use the coded data of the intra-frame coded screen every n (n ≧ 1) screens. At this time, different speed reproduction can be performed by skipping and reproducing at different intervals.

(6) The same image sequence (program) and k (k ≧ 2) types of coded data having different average bit rates, resolutions, and frame rates are stored on the storage medium. By doing so, a moving image having a desired average bit rate or resolution or frame rate can be selected and reproduced.

[0026]

Embodiments of the present invention will be described below. In the case of (1), normal reproduction is performed by using the encoded data shown in (1) to (6) in FIG. In the case of (2), 5 × speed forward playback is performed. In the case of (3), 20 × speed forward reproduction is performed. In the case of (4), playback is performed in the reverse direction at 5 times speed. In the case of (5), playback is performed in the reverse direction at 20 times speed. In the case of (6), a high-resolution still image is played back.

The normal reproduction data shown in (1) does not differ from the image data structure on the conventional normal storage medium. If there is a normal playback request, the normal playback data shown in (1)
When special playback is requested, the fast forward playback data shown in (2) and (3), the fast rewind playback data shown in (4) and (5), or the still image data shown in (6). By sequentially reading each of these, normal reproduction and special reproduction of each can be performed. However, it is necessary to randomly access and search the first access position.

In this case, only the data for special reproduction,
Since it is necessary to have extra data, the amount of accumulated data increases. However, the data for high-speed reproduction has a fast-moving screen and many scene changes, so that visually high image quality as in normal reproduction is not required. Therefore,
The data for special reproduction may be data having a high compression rate by lowering the resolution or coarsening the quantization step size of the quantizer, as compared with the data for normal reproduction. Therefore, for example, the amount of data for 20 × speed reproduction is smaller than 1/20 of the amount of data for normal reproduction.

FIG. 3 shows an example of the configuration of a coding system for high-efficiency coding of image data. Intra-frame coding is performed periodically by the motion compensation prediction + DCT coding method. The example of the inter-frame coding system is shown.

At the time of interframe coding, the switcher 11 takes out one frame of image data accumulated in the frame memory (FM) 12 via the variable delayer (VDL) 13, and the subtractor 14 inputs the input image. The result of taking the difference from the data is the discrete cosine transformer (DCT)
After performing a discrete cosine transform by 15, a quantizer (Q) 16 performs a quantization process, a variable-length coding unit 17 converts the data into variable-length coded data, and a buffer memory (BM) 18 , Output as a bitstream.

Further, the inverse quantizer (IQ) 19 performs inverse quantization processing on the quantized data,
After performing the inverse discrete cosine transformation in the inverse discrete cosine transformer (IDCT) 20, the image data for one frame from the switcher 11 is added with the data fixedly delayed by the delayer (Delay) 21 in the adder 22. And stores it in the frame memory (FM) 12.

At this time, in the motion compensation prediction unit (MC unit) 23, from the input data and the image data accumulated in the frame memory (FM) 12, the motion compensation prediction unit (MC) 2
4, the motion vector information is detected, and this motion vector information is converted into variable length coded data in the variable length coding unit 17.

Further, the activity calculation unit (Acti)
feed-forward control is performed based on the result of the property determination of the picture in the (bity) 25, and feedback control is performed based on the data accumulated in the buffer memory (BM) 18, thereby the control unit (Cont).
At 26, the quantizer (Q) 16 quantization step size is determined.

On the other hand, when performing intra-frame coding,
By inputting “0” by the switcher 11, the discrete cosine transformer (DCT) 15 performs the discrete cosine transform on the input image data. By performing such a switching periodically, the I-picture (Intra Co
Decoded picture) is inserted every fixed period (15 frames), and MPEG coding is performed.

FIG. 4 shows an I-picture for special reproduction.
Fig. 2 shows the structure of image data when using, where (1) is normal playback data, (2) is approximately 7x forward playback data, (3) is approximately 21x forward playback data, 4) is about 7 times reverse playback data, and (5) is about 21 times reverse playback data.

In FIG. 4, numerals indicate frame numbers, I is I-picture and P is P-picture.
(Predictive Coded picture), B is B-picture
e (Bidirectional Predictive Coded picture). Here, P-picture is I-picture
The difference from e is encoded every 3 frames, and B-picture is the immediately preceding I or P-pic.
The difference between the “ture” and the immediately following I or P-picture is encoded.

As shown in FIG. 4, a new file is constructed for special reproduction only by the I-picture. In the case of special reproduction, each special reproduction is performed by sequentially accessing this file. Also, if the I-picture is skipped and reproduced properly, reproduction at different speeds is possible.
At this time, by inserting the stuff bit, etc., each I-p
If the amount of information in the picture is kept constant, access can be made at regular intervals, so that different double speed reproduction can be easily realized.

FIG. 5 shows an example of accumulating a plurality of data at different bit rates and resolutions for special reproduction.
(1) is the low resolution first example (Low-1), (2) is the low resolution second example (Low-2), and (3) is the high resolution first example.
(High-1) and (4) are high-resolution second examples (H
Igh-2).

Low-1 shown in (1) has an average bit rate of 1.5 Mb / s, and the number of pixels in one frame is 3
It is 52 × 240. In Low-2 shown in (2), the average bit rate is 3.0 Mb / s, and the number of pixels in one frame is 352 × 240. High-1 shown in (3)
Has an average bit rate of 3.0 Mb / s, and the number of pixels in one frame is 720 × 480. H shown in (4)
The high-2 has an average bit rate of 9.0 Mb / s, and the number of pixels in one frame is 720 × 480.

As described above, the same image sequence (program) is individually stored as a plurality of data having a plurality of bit rates and resolutions, and the data of the bit rate and the resolution is selected according to the demand of image quality. The same image sequence (program) can be reproduced with different image quality by accessing the same.

The form of the special reproduction MPEG compressed data that can be realized by the method of the present invention will be described below. FIG.
FIG. 8 is a diagram (1) to (3) showing special playback functions, each function having the illustrated contents, and consisting of five kinds of bit streams (1) to (5) shown in the drawing. There is.

FIG. 9 illustrates the structure of the MPEG bit stream. (0) shows the source image before compression, and (1) to (5) are shown in FIGS. 6 to 8, respectively. (1)
Corresponding to the bit stream of ~ (5), (1) is data for normal playback, (2) is data for reverse playback, (3) is data for double speed forward playback, and (4) is double speed reverse playback. Data, (5) is high-speed playback data. In the figure, #XX indicates a frame number, which is I-picture, P-picture, B-.
Display the picture as shown. Also, GOP
(Groop Of Pitures) indicates the insertion interval of I-picture.

6 to 8 and 9, (1) to
The bit stream shown in (5) consists of the following data. The data for normal order reproduction shown in (1) is IPB format data that is normally MPEG-compressed. The reverse-order reproduction data shown in (2) is the same IPB format data as the normal-order reproduction data, which is obtained by compressing a frame in reverse.

The double-speed normal forward reproduction data shown in (3) is IPB format data obtained by MPEG-compressing only even frames. The double speed reverse order reproduction data shown in (4) is the same IPB format data as the double speed normal order reproduction data in which only even frames are MPEG compressed in reverse order.

The data for high speed reproduction shown in (5) is I-pi.
This is a collection of only even-numbered frames of data encoded n times in the picture. In the case of double speed reproduction, frame data is thinned out according to the double speed and sent to the client. In the reverse order, the data is sent in the reverse order. When the texture is complicated, the image quality is degraded (looks like a mosaic image), but in most cases, reproduction at 3 × speed or more seems to be the purpose of search, so there is no problem. Conceivable.

FIG. 10 shows the ratio of the disk capacity when the MPEG bit stream is realized, and (1) to (5) in the figure show (1) to (5) shown in FIG.
Corresponding to the bit stream of (5), (1) is data for forward playback, (2) is data for backward playback, (3) is data for forward normal playback, and (4) is for reverse playback at double speed. Data, (5) is high-speed playback data.

The bitstreams (1) to (5) can be created by an ordinary encoder. In the above example, on the video tape recorder (VTR), double speed data, reverse order data, and double speed reverse order data are prepared in advance by VTR editing, and the normal encoder (2)-
Create the data in (4). The data in (5) is I-pic.
It can be created simply by encoding all frames in the true.

Hereinafter, a method of generating (encoding) these five kinds of bit stream data will be described. It is assumed that this encoding board can input data from an analog VTR (NTSC system) and a digital VTR (D1 system).

The data of (1) is created by encoding online. The data of (2) is created by putting the reverse order data in advance in the D1 method VTR and encoding it offline. (3)
Data is created by putting double-speed data in a D1 system VTR in advance and encoding it offline.

The data of (4) is off-line and previously set to D1.
It is created by putting double speed data in the reverse order in the system VTR and encoding it. For the data in (5), after performing I-picture encoding online, insert a dummy bit in the short data so that the data length of each frame will be equal while picking up only even frames offline. And adjust it so that the length is the same.

FIG. 11 shows an example of a special reproduction request. In the figure, 31 is a hard disk (HD)
And a playback device and a monitor, which are placed on the client side, which are placed on the server side, are connected to each other via a communication line 33, and respond to a request from the client side. In response, the image data is sent from the server side.

FIG. 12 shows an example of a data management table. (1) is a management file for normal order data (movie A), and (2) is for double speed normal order data (movie A). It is a management file.

An example of the trick play method using the data management table will be described below. It is assumed that the normal data and the special reproduction data are managed by the management table as shown in FIG. In the management table, the frame number of the image data and the logical block number of the corresponding hard disk are written in time order.

Here, the bit stream may be written on one hard disk or striped on a plurality of hard disks. Further, the logical block number indicates not only the logical block number in the hard disk but also which hard disk is written.

Now, in FIGS. 11 and 12, time t0
Then there is a request from the client side to play movie A in the normal order,
The processing on the server side in this case will be described assuming that there is a request for normal speed reproduction at double speed at time tn and a request for normal order reproduction again at time tk.

First, at the time t0, a request for normal order reproduction of the movie A is made. Therefore, the normal order reproduction management table is read, and data is transmitted from the data of the logical block number 0 corresponding to the first frame number 0. To start.

Next, the time tn when the frame n was reproduced
Then, there was a request from the client side for double speed normal playback, so read the management table for double speed normal playback,
Data is transmitted from the data of the logical block number 100498 corresponding to the frame number n.

Next, at the time tk during which the frame k was reproduced at the normal speed of double speed, the client side again requested the normal reproduction, so the management table for the normal reproduction was read.
Data is transmitted from the data of the logical block number 12048 corresponding to the frame number k.

By having such a data management table, it is possible to quickly execute any prepared special reproduction, and to use a VTR for home use on the VOD system.
It is possible to realize the special reproduction function.

[0060]

As described above, according to the present invention, at the time of reproducing image data, only the data is sequentially read from the storage medium, and the intermediate processing such as data rearrangement by the CPU is hardly required. It is possible to realize a medium storage method of moving image coded data, which can realize special reproduction such as fast forward and fast reverse.

[Brief description of drawings]

FIG. 1 is a diagram conceptually showing a system configuration of the present invention, in which (a) shows a case where a storage medium is directly accessed,
(B) shows a case where the storage medium is accessed via a communication line.

FIG. 2 is a diagram showing a structure of image data on a storage medium, where (1) is normal reproduction data, (2) is 5 × speed forward reproduction data, (3) is 20 × speed forward reproduction data, 4) is 5x reverse playback data, (5) is 20x reverse playback data,
(6) shows high resolution still image data inserted periodically.

FIG. 3 is a diagram showing a configuration example of a coding system in the case of performing high-efficiency coding of image data.

FIG. 4 is a diagram showing a structure of image data when an I-picture is used for special reproduction, where (1) is normal reproduction data, (2) is about 7 × forward reproduction data, and (3) Indicates about 21x speed forward reproduction data, (4) indicates about 7x speed reverse reproduction data, and (5) indicates about 21x speed reverse reproduction data.

FIG. 5 is a diagram showing an example of accumulating a plurality of data at different bit rates and resolutions for special reproduction, where (1) is a low resolution first example (Low-1) and (2) is a low resolution. The second example (Low-2), (3) is the high-resolution first example (High-2).
-1) and (4) show the second example of high resolution (High-2), respectively.

FIG. 6 is a diagram (1) showing a special reproduction function.

FIG. 7 is a diagram (2) showing the special reproduction function.

FIG. 8 is a diagram (3) showing the special reproduction function.

[Fig. 9] Fig. 9 is a diagram for explaining the structure of an MPEG bit stream, in which (0) shows a source image before compression, (1) is data for forward playback, (2) is data for backward playback, and (3) ) Is the data for double speed normal reproduction, (4) is the data for double speed reverse reproduction,
(5) shows high-speed playback data.

[Fig. 10] Fig. 10 is a diagram showing a ratio of disk capacities when an MPEG bit stream is realized, where (1) is data for forward playback, (2) is data for backward playback, and (3) is double speed forward order. Reproduction data, (4) shows double speed reverse order reproduction data, and (5) shows high speed reproduction data.

FIG. 11 is a diagram showing an example of a special reproduction request.

FIG. 12 is a diagram showing an example of a data management table, in which (1) is a management file for normal data (movie A), and (2)
Shows a management table for double speed normal order data (movie A).

[Explanation of symbols]

 1 storage medium 2 reproducing means 3 communication line

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location G06T 9/00 G11B 20/10 321 Z 7736-5D H04N 5/783 H 5/93 7/24 7 / 173 G06F 15/64 450 F 15/66 330 A H04N 5/93 Z 7/13 Z (72) Inventor Masahiro Sekiguchi 1015 Ueodaanaka, Nakahara-ku, Kawasaki-shi, Kanagawa Fujitsu Limited (72) Inventor Mayumi Ozora 1015 Kamiodanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture, Fujitsu Limited

Claims (6)

[Claims] 1. A system for reproducing moving images by accumulating coded data of a moving image in a storage medium and accessing the coded data directly from a reproducing means or via a communication line. It has separate encoded data for special playback and encoded data for special playback, and by accessing either data in response to a request, normal playback and special playback can be arbitrarily selected and executed. A method of storing moving image coded data in a medium, characterized in that. 2. The encoded data for special reproduction is encoded data for fast-forward reproduction, encoded data for fast-reverse reproduction, encoded data for high-resolution still image reproduction, or any combination thereof. The medium storage method of moving image coded data according to claim 1, wherein the method is a coded data consisting of 3. A plurality of types of coded data for double speed reproduction are provided as the coded data for fast forward reproduction or fast reverse reproduction, and only the coded data for double speed reproduction required is accessed in response to a request. The method for storing the encoded moving image data medium according to claim 2, wherein the desired double speed reproduction is performed. 4. The moving picture code according to claim 2, wherein the coded data for the fast-forward reproduction or the fast-reverse reproduction includes coded data that is coded at a higher compression rate than normal reproduction. Media storage method for digitalized data. 5. The moving image according to claim 2, wherein the coded data of the intra-frame coded screen every n (n ≧ 1) screens is used as the coded data for fast-forward reproduction or fast-reverse reproduction. Medium storage method for encoded data. 6. A system for reproducing moving images by accumulating coded data of a moving image in a storage medium and accessing the coded data directly from a reproducing means or via a communication line, in the storage medium, By accumulating coded data of the same image sequence (program) and having different average bit rate, resolution, or frame rate (k ≧ 2), the desired average bit rate or resolution can be obtained. Alternatively, a method of storing moving image coded data in a medium is characterized in that a moving image of a frame rate is selected and made reproducible.