JP2004229323A - Mpeg image data recorder - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable the seamless reproduction without causing the contradiction of overflow or underflow in connection of a VBV buffer in reproduction, if new MPEG image data are additionally recorded to an already recorded MPEG image data from halfway or its tail end. <P>SOLUTION: Separate from a bit stream of MPEG image data, the recorder records a VBV buffer value at the picture coding end point of one frame preceding to an I-picture in the bit stream, a VBV buffer value at the picture coding end point of one frame preceding to a P-picture, a VBV buffer value at the recording end point, and address information for showing when each of the VBV buffer values is taken in the MPEG data. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

  The present invention relates to an MPEG image data recording device that records MPEG image data, which is image data compressed by the MPEG encoding method, on a recording medium. In particular, it is an object of the present invention to provide an MPEG image data recording apparatus capable of realizing seamless reproduction when reproducing additionally recorded MPEG image data.

  First, MPEG, which is a conventional technique, will be briefly described. MPEG is described in detail in ISO-IEC11172-2 and ITU-T H.262 / ISO-IEC13818-2, so only the outline will be described here. MPEG was established in 1988 in ISO / IEC JTC1 / SC2 (International Organization for Standardization / International Electrotechnical Commission / Technical Committee 1 / Specialized Subcommittee 2, current SC29). Moving Pictures Expert Group). MPEG1 (MPEG Phase 1) is a standard for storage media of about 1.5 Mbps, and is intended for JPEG for still image coding and for moving image compression for low transfer rates of ISDN video conferences and videophones. It inherits the basic technology of H.261 (CCITT SGXV, standardized by the current ITU-T SG15) and introduces new technology for storage media. These were established in August 1993 as ISO / IEC 11172.

  MPEG is created by combining several technologies. FIG. 5 shows a conventional MPEG encoder, which will be briefly described below. The difference between the input image and the image decoded by the motion compensation predictor 1 and the input image is calculated by a differentiator 2 to reduce a time redundant part.

  The prediction direction has three modes from the past, the future, and both. These can be switched and used for each MB (macroblock) of 16 pixels × 16 pixels. The prediction direction is determined by the picture type given to the input image. A P-picture has two modes: a mode for encoding based on prediction from the past, and a mode for independently encoding the MB without performing prediction. A B-picture has four modes for prediction from the future, prediction from the past, prediction from both, and independent encoding. An I-picture independently encodes all MBs.

  In motion compensation, a motion vector is detected by half-pel accuracy by performing pattern matching for each MB for each motion area, and is predicted after shifting by the amount of motion. The motion vector has a horizontal direction and a vertical direction, and is transmitted as additional information of the MB together with an MC (Motion Compensation) mode indicating where to predict the motion vector.

  A GOP (Group Of Picture) refers to a GOP (Group Of Picture) from an I picture to a picture preceding the next I picture. In general, about 15 pictures are used as one GOP.

  The difference image is subjected to orthogonal transformation in the DCT unit 3. DCT (Discrete Cosine Transform) is an orthogonal transform that discretely transforms an integral transform using a cosine function as an integral kernel into a finite space. In MPEG, MB is divided into four and two-dimensional DCT is performed on an 8 × 8 DCT block. In general, since a video signal has many low-frequency components and few high-frequency components, when DCT is performed, coefficients concentrate on low frequencies.

  The quantized image data (DCT coefficients) are quantized by the quantizer 4. The quantization is performed by quantizing the DCT coefficients by using the value obtained by multiplying the 8x8 two-dimensional frequency, called the quantization matrix, by the visual characteristics and the value obtained by multiplying the whole by a scalar multiplication scale as the quantization value. Calculate by value. When inverse quantization is performed by the MPEG decoder (decoder), a value close to the original DCT coefficient is obtained by multiplying by the quantization value.

  The quantized data is variable-length coded by the VLC unit 5. The direct current (DC) component of the quantized value uses DPCM (differential pulse code modulation) which is one of predictive coding. In addition, Huffman coding that performs a zigzag scan of the alternating current (AC) component from the low band to the high band, assigns a zero run length and an effective coefficient value to one event, and assigns a code with a short code length from the one with a high probability of occurrence. Is performed.

  The variable-length coded data is stored in a temporary buffer 6 and output as coded data at a predetermined transfer rate. The generated code amount of the output data for each macro block is transmitted to the code amount controller 21, and the error code amount between the generated code amount and the target code amount is fed back to the quantizer 4 to quantize the data. The code amount is controlled by adjusting.

  The quantized image data is inversely quantized by an inverse quantizer 7, inverse DCT by an inverse DCT unit 8, temporarily stored in an image memory 10 via an adder 9, and then stored in a motion compensation predictor 1. Is used as a reference decoded image for calculating a difference image.

  FIG. 6 shows an MPEG decoder (decoder) that decodes the encoded data that has been MPEG-encoded as described above.

  Incoming encoded data (stream) is buffered in a buffer 11, and data from the buffer 11 is input to a VLD unit 12. The VLD unit 12 performs variable length decoding to obtain a direct current (DC) component and an alternating current (AC) component. The alternating current (AC) component data is arranged in an 8 × 8 matrix in the order of zigzag scan from the low band to the high band. This data is input to the inverse quantizer 13 and is inversely quantized by the quantization matrix. The inversely quantized data is input to the inverse DCT unit 14, subjected to inverse DCT, and output as image data (decoded data). The decoded data is temporarily stored in the image memory 16 and then used by the motion compensation predictor 17 as a reference decoded image for calculating a difference image.

  In the case of video, the encoded bit stream has a variable length code amount for each picture. This is because MPEG uses information transforms such as DCT, quantization and Huffman coding.At the same time, it is necessary to adaptively change the amount of code allocated to each picture in order to improve image quality. This is because the entropy of the encoded image itself greatly changes, for example, in some cases, the input image is encoded as it is, and in some cases, the difference image, which is the difference between the predicted images, is encoded.

  In this case, in many cases, the code amount is controlled while allocating to the entropy ratio of the image and keeping the buffer limit. The buffer manager monitors the relationship between the generated code amount and the coding rate, and sets a target code amount so as to be within a predetermined buffer. This value is fed back to the variable-length encoder and enters the code amount controller, where the quantized value set in the quantizer is increased to suppress the generated code amount, or the quantized value is reduced to reduce the generated code amount. Or make it smaller.

  When such variable-length data is encoded at a fixed transfer rate (encoding rate), if the maximum buffer amount of the decoder is set as an upper limit, data is input at a constant speed and only a predetermined value is accumulated. From MPEG, it is prescribed by MPEG to use a model that performs instantaneous decoding at a predetermined time (NTSC video signal in units of 1 / 29.97 sec), and that the buffer does not overflow or underflow. ing. If this rule (VBV buffer rule) is adhered to, the rate in the VBV buffer changes locally, but if the observation time is long, the transfer rate will be fixed. In MPEG, this is a fixed rate. Has defined.

  In the case of the fixed transfer rate, if the generated code amount is small, the buffer occupancy is stuck at the upper limit. In this case, the code amount must be increased by adding an invalid bit so as not to overflow.

  In the case of the variable transfer rate, the definition of the fixed transfer rate is extended so that when the buffer occupancy reaches the upper limit, the reading of the decoder is stopped so that the overflow does not occur in principle. ing. FIG. 7 shows such a buffer transition. Even if the generated code amount is very small, the reading of the decoder is stopped, so there is no need to insert an invalid bit as in the case of the fixed transfer rate. Therefore, encoding is performed so that only underflow does not occur.

  Also, according to the invention of JP-A-11-74799, when editing compressed data such as MPEG data recorded on a recording medium, the VBV buffer is always fixed at the editing point in order to maintain the continuity of the MPEG data. It describes a method of performing encoding in consideration of continuity, such as controlling the amount of generated codes so as to achieve the above or encoding a GOP as a closed GOP.

  Further, according to the invention of Japanese Patent Application Laid-Open No. 11-187354, the encoded data is not restricted at all, and information indicating the data extracted as the editing material among the partial sections of the data and the reproduction order thereof are given. A method is described in which information can be described and video editing can be realized on a single recording medium without changing recorded data.

  However, in the above-described conventional method, if MPEG image data is simply connected, inconsistency occurs in the connection of the VBV buffer, and overflow or underflow occurs. In the case of fixed-length coding, the value of VBV is described in picture units, and by observing the image bit stream, the start value of VBV at the start point for additional recording can be calculated. Yes, but it also had to decode some to MPEG compressed data. Further, in the case of variable-length coding, since the value of VBV is not described at all in the syntax, it is necessary to observe and calculate the generated code amount of each picture from the beginning of the compressed data. This required more circuitry and computation time.

  In the invention of Japanese Patent Laid-Open No. 11-74799, the amount of generated code is controlled so that the VBV buffer is always fixed for each GOP so that the GOP can be edited anywhere, and the GOP is encoded as a closed GOP. Coding restrictions in consideration of continuity are imposed, which is a disadvantageous factor in terms of coding efficiency.

  In the invention of JP-A-11-187354, reproduction and display are performed as if edited, but the continuity at the editing point is incomplete, and temporary stoppage such as initialization of the MPEG data decoder buffer is performed. A phenomenon could occur.

  The present invention enables seamless connection of the VBV buffers during playback without inconsistency of overflow or underflow even when new MPEG image data is additionally recorded from the middle of the already recorded MPEG image data or from the end. It is an object of the present invention to provide an MPEG image data recording device capable of performing an accurate reproduction. Furthermore, the present invention provides a method of determining the start value of VBV at the starting point when additional recording is performed without decoding already recorded MPEG image data, thereby enabling MPEG image data to be reduced in circuit size and calculation time. It is intended to provide a recording device.

Therefore, in order to solve the above problems, the present invention provides the following recording device.
(1) In an MPEG image data recording device that records MPEG image data that is image data compressed by the MPEG encoding method,
Information on each VBV buffer occupancy value for each predetermined unit in the bit stream of the MPEG image data,
Information about an address indicating at which point in the MPEG image data the information about the VBV buffer occupancy value is information about the VBV buffer occupancy value,
An MPEG image data recording apparatus, further comprising recording means for recording the MPEG image data on a recording medium.
(2) In an MPEG image data recording device that records MPEG image data that is image data compressed by the MPEG encoding method,
Information on each VBV buffer occupancy value for each GOP unit in the bit stream of the MPEG image data;
Information about an address indicating at which point in the MPEG image data the information about the VBV buffer occupancy value is information about the VBV buffer occupancy value,
An MPEG image data recording apparatus, further comprising recording means for recording the MPEG image data on a recording medium.

  As described above, according to the present invention, even when new MPEG image data is additionally recorded from the middle of the already recorded MPEG image data or from the end, the connection of the VBV buffer during playback is inconsistent with overflow or underflow. , And seamless and high-quality reproduction is enabled. Furthermore, according to the present invention, the start value of the VBV buffer occupancy value at the start point when additional recording is performed without decoding already recorded MPEG image data can be known, so that the circuit scale can be reduced and additional recording can be performed. Processing time can be reduced.

  By using the present invention, even if new MPEG image data is additionally recorded from the middle of the already recorded MPEG image data or from the end, there is no inconsistency of overflow or underflow in the connection of the VBV buffer during reproduction. , Enabling seamless playback. Further, according to the present invention, the start value of VBV at the start point when additional recording is performed without decoding already recorded MPEG image data can be known, and the circuit size can be reduced and the calculation time can be reduced.

  The concept of the present invention will be described with reference to FIG. The recording medium records MPEG image data, which is image data compressed by the MPEG encoding method. The MPEG image data is recorded as a bit stream in which a plurality of continuously reproducible data encoded and generated in one recording unit are continuously connected.

  Apart from these encoded MPEG image data bit streams, the VBV buffer value (occupation value) at the end of picture encoding one frame before the I picture in the MPEG image data bit stream and the P picture 1 VBV buffer value (occupancy value) at the end of picture encoding before a frame, VBV buffer value (occupancy value) at the end of recording, and VBV buffer occupation value at each point in the MPEG image data where each VBV buffer value is Is recorded (in this example, a relative address from the beginning of the file of the MPEG image data). FIG. 1 shows the data structure of VBV buffer information including these data.

  VBV buffer information has a hierarchical structure. First there is an entry point information structure, followed by a VBV information structure. In the entry point information structure, the number of entry point (EP) addresses is first described in 32 bits, and the EPn (n is a natural number of 1 or more) address is described in 32 bits in order. The EPn address indicates a position where EPn information (n is a natural number of 1 or more) of the VBV information structure is described, and describes a relative address from the head of the VBV buffer information. On the other hand, the VBV information structure is described in order from the EP1 information, and the content of the EP1 information describes a relative address, a PTM value, and a VBV value in order.

  The relative address in the EPn information of the VBV information structure is, as shown in FIG. 2, the end point of the picture encoding one frame before the I picture and the end point of the picture encoding one frame before the P picture in the bit stream of the MPEG image data. The relative address from the beginning of the MPEG image data at the time and at the end of recording, for example, the unit is a byte. When recorded on a disk medium, a sector or the like is used as a relative address.

  The PTM value in the EPn information of the VBV information structure is a time stamp recorded with a clock of 90 kHz or 27 MHz in the MPEG system standard (multiplexing standard). In the MPEG standard, it is called PTS (Presentation Time Stamp) or DTS (Decoding Time Stamp). Here, the DTS is recorded as time information at the end of picture encoding one frame before the I picture in the bit stream of the MPEG image data, at the end of picture encoding one frame before the P picture, and at the end of recording. . One DTS is recorded in one picture, and in the case of an NTSC video signal, it is recorded at 90 KHz clocks at an interval of 3003 clocks per picture. Therefore, when an I picture or a P picture is present for every three pictures as in the present invention and the beginning starts from 0, the PTM information is provided at intervals of 9009, 18018. Will be described.

  The VBV value in the EPn information of the VBV information structure is a virtual buffer occupancy value of a decoder specified by MPEG. It can be derived by calculation from the generated code amount for each picture of the MPEG image data and the value of the transfer rate. As shown in FIG. 3, the point at which the picture encoding of the I-picture of the compressed bit stream information is completed one frame before the I picture , Each VBV occupation value at the end of picture coding one frame before the P picture and at the end of recording (at the position of each O mark shown in FIG. 3). Alternatively, the VBVdelay value specified in MPEG at each time point is described. This value is a value converted into a time that indicates how long it takes at the transfer rate at that time up to the VBV occupation value. In the present invention, the VBV value may be any information as long as it is information on the VBV buffer occupancy value.

  In MPEG compression, encoding is basically performed using an I or P picture type in units of three frames, such as IBB and PBB. In MPEG compression, a B picture may be predicted from both directions, so in the order of the coded bit stream, data is added only at the end of picture coding one frame before the I picture and P picture of the bit stream. Is not easy. For this reason, the present invention is characterized in that VBV information is described at the end of picture coding one frame before the I picture and P picture of the bit stream.

  Next, the configuration of an embodiment of the recording apparatus according to the present invention is shown in FIG. 4, and an operation of creating VBV buffer information while encoding image data by the MPEG encoding method will be described.

  When there is no encoded data on the recording medium 31, that is, when encoding is performed for the first time, the data reading unit 32 from the recording medium 31 does not have any data. Transmit to the detector 33. Since no data exists in the VBV buffer information detector 33, the initial value set in advance in the parameter setting unit 34, that is, the VBV value is, for example, a value of 80% of the maximum value of VBV specified by MPEG, and the PTM Time stamp information is set to 0. These initial setting values are transmitted to the image encoder 35.

  In the image encoder 35, encoding is started from an initial set value. In the image encoder 35, while performing the encoding, the generated code amount, the PTM value, and the VBV value at the end of the picture encoding one frame before the I picture of the bit stream and one frame before the P picture are calculated every time. This is transmitted to the VBV buffer information generator 36. At the same time, the encoded data is transmitted to the data writing unit 37. Further, the image encoder 35 transmits the generated code amount, the PTM value, and the VBV value when the user temporarily stops or ends the image compression recording to the VBV buffer information generator 36.

  The VBV buffer information generator 36 generates VBV buffer information data having the structure shown in FIG. 1 from the input generated code amount value, PTM value, and VBV value. Alternatively, data necessary for creating the data structure is stored in a memory and stored in a predetermined format. The information created by the VBV buffer information creating unit 36 may be written in a burst at the same time that the encoded data (MPEG image data) is written to the recording medium 31 by the data writing unit 37. The information created by the VBV buffer information creating unit 36 is recorded and held in a predetermined format when the encoded data (MPEG image data) has been written, that is, after the user has temporarily stopped or ended the image compression recording. The written data may be converted into the structure of FIG. 1 and written by the data writing unit 37.

  Next, a new image is added to the encoded data (MPEG image data) already recorded together with the VBV buffer information on the recording medium 31 from a predetermined position in the encoded data (MPEG image data). The operation for recording data will be described.

  An encoded stream (MPEG image data bit stream) is already recorded on the recording medium 31. Therefore, the data reading unit 32 reads the VBV buffer information from the recording medium 31, and obtains a VBV value, a PTM value, and a relative address. When additional recording is performed from the end of the encoded stream already recorded, the last value of the existing VBV buffer information (the VBV buffer whose relative address is the furthest from the beginning of the encoded stream) Information). On the recording medium 31, the position where the new MPEG image data is additionally recorded may be a position following the recording portion of the already recorded MPEG image data or may be a completely different position.

  When performing additional writing in the middle of an already recorded encoded stream, a user interface (not shown) is used to specify from which point of the already recorded image data the additional recording is to be started. . For example, when the designation method is the position information of the relative address of the data, the VBV value and the PTM value linked to the data closest to the relative address information in the EPn information in the structure of the VBV buffer information are calculated. Used. Also, if the designation method is the time from the start time of the data or the time stamp information of the point to be additionally recorded, similarly, using the PTM value in the EPn information of the structure of the VBV buffer information, When this value is recorded with a clock of 90 KHz, the time can be obtained by multiplying the value by 1 / 90,000 second, and a position (relative address) for additional recording on the way, a VBV value, PTM values can be obtained.

These values are input to the parameter setting unit 34, and the image encoding unit 35 starts encoding from the set values. (Based on the detected VBV value, VBV buffer control is started to perform encoding.)
On the other hand, the coded data search unit 38 searches the bit stream that has already been recorded for the position where the additional recording is to be started. The search uses a relative address of data to set a pointer to a position from the beginning of the bit stream file.

  The image encoder 35 encodes the generated code amount, the PTM value, and the VBV value at the end of picture encoding one frame before the I-picture and P-picture of the bit stream, while performing encoding, each time using a VBV buffer information generator. 36. At the same time, the encoded data is transmitted to the data writing unit 37. Further, the image encoder 35 transmits the generated code amount, the PTM value, and the VBV value when the user temporarily stops or ends the image compression recording to the VBV buffer information generator 36.

  The VBV buffer information generator 36 generates VBV buffer information data having the structure shown in FIG. 1 from the input generated code amount value, PTM value, and VBV value. Alternatively, data necessary for creating the data structure is stored in a memory and stored in a predetermined format. The information created by the VBV buffer information creating unit 36 may be written in a burst at the same time that the encoded data (MPEG image data) is written to the recording medium 31 by the data writing unit 37. The information created by the VBV buffer information creating unit 36 is recorded and held in a predetermined format when the encoded data (MPEG image data) has been written, that is, after the user has temporarily stopped or ended the image compression recording. The written data may be converted into the structure of FIG. 1 and written by the data writing unit 37.

  New MPEG image data is additionally recorded from the point of additional recording in the middle of already recorded MPEG image data.

The new MPEG image data additionally recorded in the middle of the already recorded MPEG image data may be recorded so as to overwrite the part of the recording medium 31 where the MPEG image data is already recorded. Good. Further, the image data may be recorded on a part on the recording medium 31 completely different from the part on the recording medium 31 on which the MPEG image data is already recorded. (In this case, it is necessary to link the additional recording point on the already recorded MPEG image data with the new MPEG image data.)
In the above description, the new MPEG image data to be additionally recorded has been described as being recorded on the recording medium on which the MPEG image data has already been recorded, but together with the MPEG image data read from the recording medium 31, the above method is used. It may be recorded on another recording medium.

  In the above embodiment, the system is described as a single recording apparatus. However, the recording medium may be a database via a network.

FIG. 4 is a diagram showing an embodiment of a VBV buffer information structure according to the present invention. FIG. 3 is a diagram illustrating a relationship between MPEG image data and relative addresses according to the present invention. FIG. 4 is an explanatory diagram illustrating a VBV value. FIG. 1 is a block diagram showing an embodiment of an MPEG image data recording device according to the present invention. FIG. 11 is a diagram illustrating an example of a conventional MPEG encoder. FIG. 15 is a diagram illustrating an example of a conventional MPEG decoder. FIG. 3 is a diagram for explaining the concept of a VBV buffer in MPEG.

Explanation of reference numerals

REFERENCE SIGNS LIST 31 recording medium 32 data reading unit 33 VBV buffer information detector 34 parameter setting unit 35 image encoder 36 VBV buffer information creator 37 data writing unit 38 encoded data search unit

Claims (2)

In an MPEG image data recording device that records MPEG image data that is image data compressed by the MPEG encoding method,
Information on each VBV buffer occupancy value for each predetermined unit in the bit stream of the MPEG image data,
Information about an address indicating at which point in the MPEG image data the information about the VBV buffer occupancy value is information about the VBV buffer occupancy value,
An MPEG image data recording apparatus, further comprising recording means for recording the MPEG image data on a recording medium. In an MPEG image data recording device that records MPEG image data that is image data compressed by the MPEG encoding method,
Information on each VBV buffer occupancy value for each GOP unit in the bit stream of the MPEG image data;
Information about an address indicating at which point in the MPEG image data the information about the VBV buffer occupancy value is information about the VBV buffer occupancy value,
An MPEG image data recording apparatus, further comprising recording means for recording the MPEG image data on a recording medium.

Images