JP2000152236A - Moving picture coder, multiplexing method, its system and recording and reproducing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To simplify the processing to retrieve a start code. SOLUTION: A multiplexer/demultiplexer 807 demultiplexes/multiplexes a video stream so that a start code is placed at the head of packets. An image coding/decoding device 805 inserts a stuffing byte directly before the start code, encodes the video stream so that the start position of the start code comes to the head of an n-byte period (n is an optional integer being 2 or over), manages the position of the start code in a transmission buffer in the image coder and outputs information of the start code position in the bit stream to the multiplexer/demultiplexer 807.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a moving picture encoding apparatus for encoding a digital moving picture signal with high efficiency, and to multiplex a video stream encoded thereby with an audio stream encoded with an accompanying audio signal. The present invention relates to a multiplexing method, an apparatus therefor, and a recording / reproducing apparatus.

[0002]

2. Description of the Related Art As a typical method for compressing and encoding moving images, "ITU-T White Book, Audiovisual / Multimedia-Related (H Series) Recommendations"
(Japan ITU Association, issued February 18, 1995)
The moving picture compression standard H. specified in (Document 1). 262
(Commonly known as the MPEG2 standard). The MPEG2 standard defines MPEG2 video, MPEG2 audio and MPE.
It is subdivided into a G2 system.

[0003] MPEG2 video is a standard for encoding moving images, and MPEG2 audio is a standard for encoding audio. The MPEG2 system standard is a bit stream (hereinafter, referred to as a bit stream) generated according to the MPEG2 video standard.
This is a standard for multiplexing a video stream) and a bit stream (hereinafter, referred to as an audio stream) generated according to the MPEG2 audio standard.

[0004] There are two types of MPEG2 system standards, a program stream and a transport stream. The program stream is a standard for multiplexing the image and audio of a single program, and is mainly intended for storage media. The transport stream is a standard for multiplexing images and sounds of a plurality of programs, and is mainly intended for a communication path.

[0005] A moving picture and its accompanying audio are stored in MPEG2 format.
When encoding according to the standard, it is necessary to multiplex bit streams separately encoded according to the MPEG2 video standard and the MPEG2 audio standard according to the MPEG2 system standard.

[0006] A conventional MPEG2 encoding device includes an image encoding device, an audio encoding device, and a multiplexing device. The image coding apparatus converts the input digital video signal into an MPE
Encode according to the G2 video standard. The generated video stream is output to the multiplexer in response to an output request from the multiplexer.

[0007] The audio encoding device encodes the input digital audio signal in accordance with the MPEG2 audio standard. The generated audio stream is output to the multiplexer in response to an output request from the multiplexer.

The multiplexer multiplexes the video stream and the audio stream according to the MPEG2 system standard and outputs the multiplexed bit stream.

Conversely, when decoding and reproducing a bit stream multiplexed according to the MPEG2 system standard, the multiplexed bit stream is separated into a video stream and an audio stream according to the MPEG2 system standard, and the respective streams are separated. Is decoded and reproduced according to the MPEG2 video standard and the MPEG2 audio standard.

[0010]

A code called a start code is provided in the MPEG2 video standard. This start code is a code that is determined not to be generated by a combination of other codes, and is a code that can be immediately distinguished from the other codes. This start code includes
sequence_header_code or grou
p_start_code or picture_star
There are t_code, etc., which are all 4-byte codes.
These start codes are respectively included at the head of a code group called a sequence header, a group of picture header, or a picture header. These start codes need to be able to be searched at high speed from the bit stream for the following reasons.

According to the MPEG2 standard, GOP is used to support trick modes such as fast forward and rewind during playback of stored media, playback from the middle and reverse playback.
(Group of picture) structure is adopted. GOP
Is a screen group structure that always includes an I picture (a screen coded only with information in one screen), and enables random access in GOP units. The standard specifies that a sequence header can be added immediately before the data of each GOP for random access required for trick mode reproduction. The sequence header added to each GOP is used for cueing for random access in trick mode.

The following procedure is required to decode a bit stream of the MPEG2 system standard recorded on a storage medium and perform trick mode reproduction. First, the bit stream read from the storage medium is stored in an MPEG format.
According to the two-system standard, a video stream and an audio stream are separated. Then, the video stream is analyzed to search for a sequence_header_code which is a start code included in the sequence header. Next, the GOP data following the sequence header is decoded and reproduced. In the above procedure, the processing time required until the start code is searched and found is a delay time that cannot be ignored in performing the trick mode reproduction.

Thus, in the MPEG2 video standard,
In order to cope with random access, a sequence header is added to each GOP. However, it is not always necessary to add a sequence header to each GOP, and often it is not added in consideration of information efficiency. Even in this case, some functions of the trick mode can be realized by using a group of picture header, which is a code indicating a GOP delimiter, for cueing.

To perform trick mode reproduction by utilizing this, the following procedure is required. First, a bit stream is separated into a video stream and an audio stream according to the MPEG2 system standard. Then, the video stream is analyzed, and group_start which is a start code included in the group of picture header is analyzed.
Search for _code. Next, the GOP data following the group of picture header is decoded and reproduced. In the above procedure, the processing time required until the start code is searched and found is a delay time that cannot be ignored in performing the trick mode reproduction.

[0015] One of the techniques for fast-forwarding playback of a bit stream of the MPEG2 standard is to skip B pictures and
There is a method of reproducing only a picture and a P picture.

The following procedure is required to decode a bit stream of the MPEG2 system standard recorded on a storage medium and perform fast-forward reproduction of an image. First, a bit stream is separated into a video stream and an audio stream according to the MPEG2 system standard. Then, the video stream is analyzed to search for a picture_start_code which is a start code of the picture header. Next, it is determined from the picture type recorded in the picture header whether the picture is an I picture, a P picture, or a B picture, and the decoding / reproducing process is performed only for the I picture and the P picture. If the picture is a B picture, pi which is the start code of the next picture header
Search for "cure_start_code". In the above procedure, the processing time required for searching for and finding a start code is a delay time that cannot be ignored in performing fast-forward playback.

For the reasons described above, when decoding and reproducing an encoded bit stream recorded on a storage medium, it is necessary to reduce the processing time for searching for a start code.

Accordingly, a first problem to be solved by the present invention is to reduce the processing time for searching for a start code.

The MPEG2 system standard employs a packet multiplexing method used in many time division multiplexing methods. The program stream of the MPEG2 system is a system in which a video stream and an audio stream are divided into bit streams of appropriate length called packets, and packets of the video stream and packets of the audio stream are transmitted in a time-division manner. To these packets, a code called a packet header is added to the head for identifying the attribute of video or audio. In order to synchronize the image with the audio, information called a time stamp is appropriately added to the packet header for each unit of decoding and reproduction.

The unit for decoding and reproducing MPEG2 video is one frame, and the unit for decoding and reproducing MPEG2 audio is one audio frame. A code called a picture header is added to the video stream immediately before the data of the frame as a code indicating a delimiter for each frame. When a packet including the head of the picture header exists in a certain pack in the pit stream, time stamp information must be added to the packet header.

In the conventional MPEG2 encoding apparatus, a multiplexing apparatus analyzes a video stream and detects a picture_st, which is a start code of a picture header, in order to detect a break for each frame in the video stream.
Art_code had to be detected. In the video stream, since the code amount for each frame is variable, the period of one frame is not constant. The position where the picture header serving as a frame delimiter appears in the bit stream is unknown to the multiplexer. Therefore,
The multiplexer must perform start code detection processing on the entire range of the input bit stream. Reducing the amount of start code detection processing to be performed by the multiplexer is a second problem to be solved by the present invention.

Further, a third object of the present invention is to improve the performance of the recording / reproducing apparatus and to reduce the size or reduce the power consumption.

[0023]

The first object of the present invention is to divide a bit stream in a multiplexing process according to the MPEG system so that each start code is at the head of a packet, and , A packet including a start code is to be divided so as to be at least 4 bytes, which is the code length of each start code.

With such division, since the start code is always contained in one packet, the start code can be retrieved from each packet alone without combining the packets of the video stream. If the packet is not divided in this manner, one start code may be divided into two or more packets. Therefore, the start code must be searched while joining the preceding and succeeding packets.

If the start code is located at the beginning of the packet, the start of each packet is set to the comparison position and the start code is set to 1
The start code can be determined only by performing the comparison operation twice. If the start code is not located at the beginning of the packet, it is necessary to repeat the comparison operation while shifting the comparison start position.

By these two effects, the processing time for searching for a start code from a bit stream can be greatly reduced.

Means for solving the second problem of the present invention will be described. The start position of the start code in the video stream is defined by the MPEG2 video standard to be located at a position of one byte cycle from the beginning of the bit stream. To this end, the multiplexing apparatus transmits a picture_s which is a start code of the picture header.
In order to perform start_code detection processing, it is necessary to compare the start position with the start code while updating the comparison position of the video stream for each byte. On the other hand, M
According to the PEG2 video standard, a code word called a stuffing byte is allowed to be inserted immediately before every start code. A stuffing byte is a codeword that can be inserted into a bitstream and is discarded in decoding processing. The stuffing byte is a code word consisting of one byte, and an arbitrary number of stuffing bytes can be inserted. By inserting the stuffing bytes, the code length of the bit stream can be increased in byte units.

According to the present invention, in an image encoding apparatus, when encoding an image signal, a stuffing byte is inserted immediately before a start code, and the start position of the start code is set to an n-byte cycle (n is (An arbitrary integer of 2 or more). As a result, the multiplexing device has only to perform the comparison process while updating the comparison position between the bit stream and the start code every n bytes (n is an arbitrary integer of 2 or more), and reduces the detection processing amount by one. / N.

Further, the bit width of the bit stream signal output from the image encoding device to the multiplexing device is set to 4 bytes (32 bits) which is the code length of the start code. Then, the encoding is performed so that the start position of the start code is also at the beginning of the 4-byte cycle. By performing such a process, the start code is exactly contained in the 4-byte bit stream signal input at a time by the multiplexer. Without such treatment,
Since the start code is not always contained in the bit stream signal input at a time by the multiplexer, while connecting the bit stream signals input before and after,
The comparison with the start code must be performed by shifting the comparison start position every byte.

Therefore, by performing such processing, the presence / absence of a start code can be determined by comparing the input bit stream signal with the start code once, and the start code detection processing can be performed. The amount can be greatly reduced.

Next, another means for solving the second problem will be described. The image coding device manages the position of the start code in the transmission buffer in the image coding device. This image encoding device outputs not only the bit stream but also information on the position of the start code in the bit stream to the multiplexer. There are the following two methods as a specific method for managing the start code in the transmission buffer and outputting the information on the position of the start code.

The first method is a method in which the image encoding device outputs the number of bit stream reading operations to be performed until the multiplexing device reads the next start code. By storing the address of the start code in the transmission buffer and comparing it with the address being read from the transmission buffer, it is possible to calculate how many times the start code is read in the future.
By obtaining this information, the multiplexing device can determine at least whether or not the head of the start code is included in the bit stream obtained by one read operation. Therefore, the search range of the start code can be limited to the bit width of the bit stream obtained by one read operation. Furthermore, if the above method is used in combination and the image coding apparatus performs stuffing so that the start position of the start code is the head position of one read operation, the multiplexing apparatus completely determines the position of the start code. As a result, the start code detection process is completely unnecessary. Thus, the second problem can be solved.

The second method is a method of dividing the transmission buffer in the image encoding device into a plurality of regions and outputting a signal indicating whether or not a start code is included in each of the divided regions. By obtaining this information, the multiplexing apparatus can narrow the search range of the start code to the range of the divided areas. The second method has an advantage over the first method that the multiplexer can know not only the position of the next start code but also the positions of all start codes existing in the transmission buffer. Have.

A third object of the present invention is to realize a high-performance, small-sized or low-power-consumption recording / reproducing apparatus by employing the above-described means in an encoding apparatus or a multiplexing apparatus.

[0035]

Embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing an example of the MPEG2 encoding apparatus. This encoding device includes an image input terminal 101, an audio input terminal 102, an image encoding device 10
3, a speech encoding device 104, a multiplexing device 105,
A bit stream output terminal 106 is provided.

The image encoding device 103 has an image input terminal 1
01 from the digital moving image signal 107
Encode according to the EG2 video standard. The generated video stream 109 is output to the multiplexer 105 in response to an output request 110 from the multiplexer 105.

The speech encoding device 104 has a speech input terminal 1
02 from the digital audio signal 108
Encode according to the G2 audio standard. The generated audio stream 111 is output to the multiplexer 105 in response to an output request 112 from the multiplexer 105.

The multiplexing device 105 controls the video stream 1
09 and the audio stream 111 are multiplexed according to the MPEG2 system standard. The multiplexed bit stream 113 is output to the bit stream output terminal 106.

Conversely, when decoding and reproducing a bit stream multiplexed according to the MPEG2 system standard, the multiplexed bit stream is separated into a video stream and an audio stream according to the MPEG2 system standard, and the respective streams are separated. Is decoded and reproduced according to the MPEG2 video standard and the MPEG2 audio standard.

FIG. 2 is a block diagram showing a first embodiment of the present invention.
FIG. 3 is a block diagram showing in detail a multiplexing device 105 in the PEG2 encoding device. The first embodiment is an embodiment for solving the first problem, and the multiplexing device 105 is devised.

The multiplexing device 105 in the MPEG2 encoding device includes a video stream dividing unit 201, a start code detecting unit 202, and a header detecting unit 203.
, An audio stream dividing unit 204, a packet multiplexing control unit 205, and a packet selector 206.

The start code detecting means 202 outputs the video stream 109 output from the image encoding device 103.
Start code (sequence_he)
ader_code and group_start_code
e and picture_start_code). The start code detection signal 208 is output to the video stream dividing means 201 and the packet multiplexing control means 205.

The video stream dividing means 201 refers to the start code detection signal 208. When the start code is not detected, the video stream
Is divided by a fixed code length. When the start code is detected, the video stream is divided immediately before the start code and the packet is cut off even if the code length is shorter than the predetermined code length. Then, the video stream dividing means 2
01 generates a packet 207 by adding a packet header to the head of the divided video stream, and outputs the packet 207 to the packet selector 206. Start code (picture
_Start_code), time stamp information is added to the packet header.

The header detector 203 detects a header added to each audio frame from the audio stream 111 output from the audio encoder 104. The header detection signal 209 is output to the audio stream dividing means 204 and the packet multiplexing control means 205.

The audio stream dividing means 204 comprises:
The audio stream 111 is divided by a fixed code length, and a packet 210 is generated by adding a packet header to the beginning,
Output to the packet selector 206. Further, the audio stream dividing means 204 outputs the header detection signal 2
Referring to FIG. 09, if a header is detected, time stamp information is added to the packet header.

The packet multiplexing control means 205 determines the output order of the video stream packet 207 and the audio stream packet 210 based on the information of the start code detection signal 208 and the header detection signal 209.
The control information 205 is output to the packet selector 206.

The packet selector 206 selects a video stream packet 207 and an audio stream packet 210 in accordance with a control signal 211 from the packet multiplexing control means 205, and outputs a multiplexed bit stream 113 by adding a pack header. .

FIG. 3 shows a method of dividing a video stream according to the first embodiment.

(A) is a video stream before division. Here, it is assumed that the video stream before division has a code length of 32 bytes, and a start code exists in a 4-byte section from the 15th byte to the 18th byte.

(B) is a method of dividing a video stream by a conventional multiplexing method. In the conventional method, the video stream is divided regardless of the position of the start code. In this example, the packet is divided so that one packet has 8 bytes. A 32-byte video stream is divided into four packets. First byte and second of start code
The byte is included in packet 2 and the third and fourth bytes are included in packet 3.

(C) is a method for dividing a video stream by the multiplexing method of the present invention. Even in this method, the packet is basically divided so that one packet becomes 8 bytes. However, if the start code is placed in a portion other than the beginning of the packet, the packet is aborted with the data immediately before the start code even if it is less than 8 bytes. As a result of such division, the video stream is divided from packet A to packet E.
Since the first 8 bytes of the video stream do not include a start code, it is used as packet A as it is. Next 8
Since the bytes include the first two bytes of the start code at the end, the six bytes excluding the start code portion are used as the packet B. The next 8 bytes including the start code are used as a packet C. Similarly, the next 8 bytes are set to packet D, and the remaining 2 bytes are set to packet E.

The effectiveness of the above dividing method will be described with reference to FIG. FIG. 4 is a diagram showing a procedure for searching for a start code from a packet.

FIG. 3A shows a packet sequence generated by the conventional dividing method, which is the same as FIG. 3B.
In the conventional division method, since the start code may be divided into a plurality of packets, these packets must be combined once and the video stream must be restored as shown in (b).

Then, as shown in (c), the video stream and the start code are compared. In comparison 1 in (c), the first 4 bytes of the video stream are compared with the start code, and it is determined that they do not match.
(C) In comparison 2, the comparison start position is advanced by one byte and compared with the start code, and it is determined that they do not match. Less than,
Similarly, the comparison with the start code is advanced while the comparison start position is advanced one byte at a time. That is, in this case, 15 comparison operations are required until the start code is found.

(A ′) is a packet sequence generated by the division method according to the present invention, which is the same packet as FIG. 3 (c). In the division method according to the present invention, since the start code is always contained in one packet, there is no need to combine the packets and restore the video stream as in the conventional case, and immediately compare the packet with the start code. Can be. As shown in (c '), the packet is compared with the start code. In comparison 1 ', the first 4 bytes of packet A are compared with the start code,
Judge as mismatch. Since the start code is always located at the head of the packet, it is determined that the packet A does not include the start code. Therefore, it is possible to immediately proceed to the comparison with the packet B. In comparison 2 '
The first 4 bytes of the packet B are compared with the start code, and it is determined that they do not match. Next, in comparison 3 ', packet C
Is compared with the start code of 4 bytes, and it is determined that they match. In this case, the comparison operation until the start code is found is three times.

According to the present invention, a start code can be directly searched from a packet without reconstructing a video stream by combining a packet sequence, and only the head of each packet is compared for start code detection. Since the process only needs to be performed, the start code can be searched at high speed with a smaller processing amount than the conventional method.

FIG. 5 shows a second embodiment of the present invention.
FIG. 2 is a block diagram showing in detail an image encoding device 103 in the PEG2 encoding device. The second embodiment is an embodiment for solving the second problem, and the image encoding device 103 is devised.

The image encoding device 103 in the MPEG2 encoding device is configured to adjust the bit width of the video stream 109 to 4 bytes (32 bits) which is the code length of the start code.

The image encoding device 103 includes an encoding unit 30
1, a stuffing byte insertion unit 302, and a transmission buffer 303. The difference from the conventional image coding apparatus is that a stuffing byte inserting means 302 is provided.

The encoding means 301 includes an image input terminal 101
The digital video signal 107 input from the
2 Encode according to the video standard. The generated video stream is temporarily stored in the transmission buffer 303 through the stuffing byte insertion unit 302.

The transmission buffer 303 is provided for the multiplexer 105
Video stream 10 in response to an output request 110 from
9 is output to the multiplexer 105.

The encoding means 301 supplies the stuffing byte insertion means 302 with pic as one of the start codes.
When outputting the tue_start_code, an enable signal is sent to the stuffing byte inserting means 302.

The stuffing byte insertion means 302
When the enable signal is not sent from the encoding unit 301, the video stream input from the encoding unit 301 is output to the multiplexing device 105 as it is. Then, the stuffing byte insertion means 302
Manages to which phase of the 4-byte period the bit stream output from the transmission buffer 3 is transmitted.
The stuffing byte is output to 03 at the end of the 4-byte cycle.

FIG. 6 shows the bit stream 109 output from the transmission buffer 303 near the picture header. In FIG. 6, (a), (b),
(C) and (d) are bit streams in the case of a conventional image encoding device. These (a), (b), (c),
(D) is a continuous series of bit streams,
(A), (b), (c), and (d) are output in this order. In this example, the picture header of the I picture is included.
If no special extension is performed, the code length of the picture header of the I picture is 8 bytes according to the MPEG2 standard. The first four bytes of the picture header are occupied by a start code. The start code is located at the third and fourth bytes of (a) and at the first and second bytes of (b).

(A ′), (b ′),
(C ′) and (d ′) are bit streams when the present invention is applied. (A '), (b'), (c '),
(D ') is a continuous series of bit streams,
(A '), (b'), (c '), and (d') are output in this order.

In this embodiment, the stuffing byte insertion means 302 inserts two bytes of stuffing bytes immediately before the start code so that the start position of the start code is at the beginning of a 4-byte unit. In the case of (a), the start code immediately follows from the third byte immediately after the data of the previous picture, but in (a '), the stuffing bytes are output in the third and fourth bytes. . As a result, the starting position in 4-byte units,
That is, the head of the start code is located at the first byte of (b ').

In the output of the conventional bit stream, the start code is divided into two outputs. On the other hand, in the case of using the present invention, the start code fits in one bit stream output. I have. Therefore, in the conventional apparatus, the multiplexing apparatus 105 must perform the comparison process for detecting the start code by shifting the comparison start position every byte while connecting the bit streams input before and after.

On the other hand, in the case of the present invention, the presence / absence of the start code can be determined by comparing the 4-byte bit stream input at a time with the start code once. The amount of detection processing can be significantly reduced.

FIG. 7 shows a third embodiment M of the present invention.
It is a block diagram which shows the image encoding device 103 in a PEG2 encoding device in detail. This third embodiment is another embodiment of the image encoding device 103.

The image coding apparatus 10 according to this embodiment
3 includes an encoding unit 301, a transmission buffer 303, a start code address management unit 504, and an address comparison unit 505.

The transmission buffer 303 includes a RAM 501,
A write address management unit 502 and a read address management unit 503 are provided. The transmission buffer 303 is a so-called ring buffer, and advances the write address 507 for the RAM 501 every time a bit stream is input. When the write address 507 advances to the maximum address of the RAM 501, it returns to the minimum address and continues the operation. .

The encoding means 301 includes the image input terminal 101
The digital video signal 107 input from the
2 Encode according to the video standard. The generated video stream is temporarily stored in the RAM 501.

The encoding means 301 outputs an enable signal 506 to the write address management means 502 when outputting the bit stream to the RAM 501. The write address management means 502
1 in response to an enable signal 506 output from
Update the write address 507 for M501. R
The AM 501 writes the bit stream output from the encoding unit 301 to the write address 507 output from the write address management unit 502.

The read address management means 503 responds to the read request 110 from the
Update the read address for M501. RAM
501 outputs the bit stream 109 stored in the read address 508 output from the read address management means 503 to the multiplexer 105.

The encoding means 301 stores a picture_picture, one of the start codes, in the transmission buffer 303.
When outputting a bit stream including the start of the start_code, an enable signal 506 is output to the write address management means 502 and an enable signal 509 is also output to the start code address management means 504.

Start code address management means 504
Has a register having a bit width capable of storing the address of the RAM 501, and upon receiving the enable signal 509, stores a write address 507 output from the write address management means 502. The start code address management means 504 stores the address 510 of the stored start code.
Is output to the address comparison means 505.

The address comparing means 505 compares the address 510 in which the start code is stored with the address 508 which is currently being read, and determines how many times the read operation will read the bit stream including the start of the start code. calculate. The address comparing means 505 outputs to the multiplexer 105 a read operation count information signal 511 to be performed until a bit stream including the start of the start code is read.

A plurality of start codes may be stored in the RAM 501. Therefore, the start code address management means 504 needs to have a plurality of registers for storing the address where the start code is written. The number of registers that the start code address management means 504 should have is determined by the capacity of the RAM 501 and the type of start code to be managed. For example, assuming that the unit of reading / writing / storage is 32 bits (4 bytes), has a 6-bit address space (64 addresses), and has a capacity of 256 bytes, the minimum size of one picture allowed by the MPEG2 standard is assumed. Is 25 bytes, and in this case, a maximum of 10 picture_start_codes are stored in the RAM 50.
1 may be stored. Therefore, the start code address management means 504 needs to have ten 6-bit registers for storing addresses. The ten registers are numbered from '0' to '9', and are used in numerical order. After register 9, register 0 is used.

The start code address management means 504 manages the number of the register holding the address to be output to the address comparison means 505, and stores the address stored in the register of that number in the start code address 5
Output as 10. Further, the start code address management means 504 manages the number of the register storing the address most recently, and upon receiving the enable signal 509, stores the address of the start code in the register next to the register storing the address most recently. I do.

The address comparing means 505 outputs the address 51
0 is compared with the currently read address 508, and when the start code indicated by the address 510 is read, a signal 512 for requesting the discard of the address 510 is output to the start code address management means 504. The start code address management means 504 stores the address 510
When the signal 512 requesting the discard of the address is received, the register number holding the address to be output to the address comparing means 505 is incremented by one.

FIG. 8 is a timing chart of an encoding process executed by the image encoding device 103 according to the third embodiment shown in FIG. (A) is a reference clock of the bit stream 109 output by the image encoding device 103, (b) is a read request signal 110 of the multiplexer 105, (c) is a bit stream 109 output by the transmission buffer 303, (D) shows the address comparing means 50
Reference numeral 5 denotes a read count information signal 511 to be performed until a start code to be output to the multiplexer 105 is read.

The bit stream 109 [FIG. 8 (c)] is output with a delay of one clock from the read request signal 110 [FIG. 8 (b)] from the multiplexer 105. The bit stream 109 includes two start codes [start code (1) and start code (2)]. In the initial state in FIG. 8, the number of times the multiplexing device 105 performs the reading operation before the start code (1) is read is five. The initial state in FIG. 8D is “5”. Then, every time the multiplexer 105 performs a reading operation, the value in FIG. 8D is decremented by '1'. For this reason, it becomes '4' at the next clock of the first read operation.
It changes to '2' and '1' and becomes '0' while the start code (1) is being output. It takes seven clocks from the end of reading the start code (1) to the start of reading the next start code (2). However, since the read request signal 110 becomes '0' halfway, Multiplexing device 10
The number of times that 5 performs the reading operation is six times. In the next clock after the reading of the start code (1) is completed, the value in FIG. 8D becomes “6”. Thereafter, every time the reading operation is performed, the value in FIG. 4 ',
It changes to '3', '2', '1' and becomes '0' while the start code (2) is being output.

The multiplexing device 105 inputs, from the image encoding device 103, information on the number of times of read operations to be performed until the start code is read, so that the start code is included in the bit stream obtained by one read operation. You can know if the head exists. When the second embodiment is combined with the third embodiment, the image coding apparatus 103 sets the start position of the start code to be the start of the bit stream obtained by one reading operation. Since the process of inserting the stuffing byte is performed, the multiplexing device 105 can completely determine the position of the start code without performing any process of detecting the start code.

FIG. 9 is a block diagram showing a fourth embodiment of the present invention.
It is a block diagram which shows the image encoding device 103 in a PEG2 encoding device in detail. The fourth embodiment is still another embodiment of the image encoding device 103.

The image encoding device 10 according to this embodiment
3 denotes an image encoding unit 301 and a transmission buffer 303
And a start code management register 701.

The transmission buffer 303 includes a RAM 501,
A write address management unit 502 and a read address management unit 503 are provided.

The capacity of the RAM 501 is 256 bytes.

The transmission buffer 303 is configured in the same ring buffer system as in the third embodiment.

The start code management register 701 stores R
The capacity of 256 bytes of the AM 501 is divided and managed in 16 areas, and a 1-bit register corresponding to each divided 16-byte area is provided.

FIG. 10 shows the start code management register 7
01 shows a register configuration. Each register stores one of the start codes, picture_star, in the corresponding 16-byte area of the transmission buffer 303.
If t_code exists, it becomes '1'; otherwise, it becomes '0'. The register corresponding to the area to which the address where the read operation is currently performed belongs to the 0th bit, and thereafter, the area is associated with the 15th bit in the order of the read operation. A 16-bit register value 702 corresponding to a total of 16 areas is output to the multiplexer 105.

The encoding means 301 includes the image input terminal 101
The digital video signal 107 input from the
Encode according to video standards. The generated video stream is temporarily stored in the RAM 501. Encoding means 30
1 outputs an enable signal 506 to the write address management means 502 at the same time as outputting the bit stream to the RAM 501.

The write address management means 502 updates the write address for the RAM 501 according to the enable signal 506 output from the encoding means 301.
The RAM 501 writes the bit stream output from the encoding unit 301 at the write address 507 output by the write address management unit 502.

When outputting a bit stream including the head of a picture start code to the RAM 501, the encoding means 301 outputs an enable signal to the write address management means 502 and also outputs an enable signal 509 to the start code management register 701. Is output. Upon receiving this enable signal 509, the start code management register 701 sets the bit corresponding to the area to which the write address output by the write address management means 502 belongs to "1".

The read address management means 503 responds to the read request 110 from the
The read address 508 for M501 is updated.
The RAM 501 outputs the bit stream stored at the read address 508 output from the read address management unit 503 to the multiplexer 105. Bit stream 1 output from RAM 501 in one read operation
09 is 4 bytes. Every time the read operation is performed four times, the area to which the address where the read operation is performed belongs moves to the next position. Therefore, the register 701 is shifted rightward by one bit.

The multiplexing device 105 includes the image encoding device 10
By referring to the output signal 702 of the register 701 that manages the presence or absence of the start code in the transmission buffer 303 in No. 3, the detection range of the start code can be limited within one area, that is, within 16 bytes. In addition, by combining the second embodiment with the fourth embodiment, the number of comparison processes required for the start code detection process is four at the maximum.

Note that the read count information signal 5 to be read before the start code is read in the third embodiment.
A signal 702 indicating the presence or absence of a start code for each of the divided buffer areas in the eleventh and fourth embodiments is output from the image encoding device 103 to the multiplexing device 105 using a signal line different from the bit stream 109. However, these signals may be multiplexed on the signal lines of the bit stream 109. For example, the multiplexer 10
5 is configured using a general-purpose CPU, and the signals 511 and 702 are held in registers that can be read from the general-purpose CPU.
The bit stream 109 and the signals 511 and 702 held in the register may be output via the PU bus.

FIG. 11 is a block diagram of a recording and reproducing apparatus according to a fifth embodiment of the present invention. This image recording / reproducing apparatus is an embodiment in which the image encoding apparatus and the multiplexing apparatus in the above-described MPEG2 encoding apparatus are used, and further, peripheral devices are combined.

This recording / reproducing apparatus includes an imaging device 801 and
An image data processing device 802, a voice input device 803,
The audio data processing device 804, the image encoding / decoding device 805, the audio encoding / decoding device 806,
It comprises a separating device 807, a recording / reproducing signal processing device 808, a recording / reproducing means 809, a video output device 810, a display device 811, an audio output device 812, and a sound generator 813.

The imaging device 801 includes a lens 801a and a CC
A D sensor 801b is provided, and converts an optical image into an electric signal to generate an analog image signal.

The image data processing device 802 processes an analog image signal to generate a digital moving image signal 107.

The voice input device 803 converts voice into an electric signal to generate an analog voice signal.

The audio data processing device 804 processes an analog audio signal to generate a digital audio signal 108.

The image encoding / decoding device 805 is configured by combining the image encoding device 103 in the above-described embodiment with a known image decoding device. At the time of recording, the digital moving image input from the image data processing device 802 is used. The signal 107 is coded according to the MPEG2 video standard, and the generated bit stream 109 is multiplexed / separated by
Output to the multiplexer / demultiplexer 807 in response to an output request from At the time of reproduction, the bit stream 109 output from the multiplexing / demultiplexing device 807 is input and the MPE
Decoding is performed according to the G2 video standard to generate a digital image signal 107 '.

The speech encoding / decoding device 806 is configured by combining the speech encoding device 104 in the above-described embodiment and a known speech decoding device. At the time of recording, the digital speech signal input from the speech data processing device 804 is used. 1
08 in accordance with the MPEG2 audio standard, and the generated bit stream 111
7 in response to an output request from the multiplexer / demultiplexer 807. At the time of reproduction, the multiplexing / separating device 807 is used.
Input the bit stream 111 output from
The digital audio signal 108 'is decoded according to the EG2 audio standard.

The multiplexing / demultiplexing device 807 is configured by combining the multiplexing device 105 of the above-described embodiment with a well-known demultiplexing device.
And audio stream 111 are multiplexed according to the MPEG2 system standard to generate a bit stream 113, which is output to a recording / playback signal processing device 808.
The multiplexed bit stream 113 output from the recording / reproducing processor 808 is converted into a video stream 109 and an audio stream 1 according to the MPEG2 system standard.
11 and input to the image encoding / decoding device 805 and the audio encoding / decoding device 806.

The recording / reproducing signal processor 808 sends the multiplexed bit stream 11 to the recording / reproducing means 809.
3 is recorded and played back.

The recording / reproducing means 809 is a recording / reproducing means using a magnetic hard disk, DVD / RAM, CD / RW or magnetic tape as a recording medium.

The video output device 810 converts the digital moving image signal 107 (107 ′) output from the image data processing device 802 or the image encoding / decoding device 805 into a video signal and supplies the video signal to the display device 811.

The display device 811 has a built-in or external L
It is a CD or CRT display.

The audio output device 812 converts the digital audio signal 108 ′ output from the audio data processing device 804 and the audio encoding / decoding device 806 into an audio signal and supplies it to the sound generator 813.

The sound generator 813 is a built-in or external speaker.

According to such a recording / reproducing apparatus, it is possible to realize a high-performance, small-sized or low-power-consumption recording / reproducing apparatus by inheriting the functions and effects obtained in each of the above-described embodiments.

In particular, when a part of the constituent means is constituted by the control function means realized by the CPU and the processing program, the shortening of the processing time and the reduction of the processing amount can reduce the load of the CPC and reduce the processing capacity. The use of the CPU and the responsibility for other control processes are enabled, which is convenient for miniaturization of the device. In addition, since the power consumption is reduced, the life of the battery of a portable device powered by the battery is shortened. Can be lengthened.

Such a recording / reproducing apparatus can be implemented as an MPEG camera, a video camera, or a VTR with high performance, small size or low power consumption by selecting and using constituent means.

[0116]

As described above, according to the present invention, by multiplexing so that the start code defined by the MPEG standard is arranged at the head of the packet,
By multiplexing the packet including the start code so that the code length of the packet is larger than the start code, it is possible to reduce the start code search time during reproduction.

Further, the moving picture encoding apparatus encodes the start position of the picture header in the bit stream so as to have a fixed byte cycle, so that the amount of picture header detection for the video stream to be performed by the multiplexing apparatus is reduced. Can be reduced.

Further, the image encoding device manages the position of the picture header in the transmission buffer in the image encoding device, and outputs not only the information of the bit header but also the position of the picture header in the bit stream to the multiplexer. By doing so, the amount of picture header detection processing for a video stream to be performed by the multiplexing device can be reduced.

Therefore, it is possible to realize a recording / reproducing apparatus with high performance, small size and low power consumption.

[Brief description of the drawings]

FIG. 1 is a block diagram of an MPEG2 encoding apparatus according to the present invention.

FIG. 2 is a block diagram illustrating in detail a multiplexing device in the MPEG2 encoding device according to the first embodiment of the present invention.

FIG. 3 is an explanatory diagram showing a video stream division method according to the first embodiment of the present invention.

FIG. 4 is an explanatory diagram showing a search procedure of a start code according to the first embodiment of the present invention.

FIG. 5 is a block diagram showing in detail an image encoding device in an MPEG2 encoding device according to a second embodiment of the present invention.

FIG. 6 is an explanatory diagram showing a bit stream near a picture header according to the second embodiment of the present invention.

FIG. 7 is a block diagram showing in detail an image encoding device in an MPEG2 encoding device according to a third embodiment of the present invention.

FIG. 8 is a timing chart of an encoding process executed by an image encoding device according to a third embodiment of the present invention.

FIG. 9 is a block diagram showing in detail an image encoding device in an MPEG2 encoding device according to a fourth embodiment of the present invention.

FIG. 10 is a diagram illustrating a configuration of a start code management register according to a fourth embodiment of the present invention.

FIG. 11 is a block diagram of a recording / reproducing apparatus according to a fifth embodiment of the present invention.

[Explanation of symbols]

101 image input terminal, 102 audio input terminal, 103
... image coding device, 104 ... voice coding device, 105 ...
Multiplexer, 106 ... bit stream output terminal, 10
7 digital video signal, 108 digital audio signal, 109 video stream, 110 video stream output request, 111 audio stream, 112
... Audio stream output request, 113 ... Multiplexed bit stream, 201 ... Video stream dividing means, 202 ... Start code detecting means, 203 ... Header detecting means, 204 ... Audio stream dividing means, 2
05: packet multiplexing control means, 206: packet selector, 207: video stream packet, 208: start code detection signal, 209: header detection signal from audio stream, 210: audio stream packet, 211: packet header, 212: detection signal of start code, 301: encoding means, 302
Stuffing byte insertion means, 303 transmission buffer, 501 RAM, 502 write address management means, 503 read address management means, 504 start code address management means, 505 address comparison means, 506 enable signal, 507 Write address, 508 ... Read address, 509 ... Enable signal, 510 ... Start code address, 511 ... Read operation count information to be performed until the start code is read out, 512 ... Start code address 510 discard request, 701 ... Start code management register, 7
02: signal indicating the presence or absence of a start code for each divided buffer area.

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Eiichi Tanaka 5-2-1, Josuihonmachi, Kodaira-shi, Tokyo Inside System LSI Development Center, Hitachi, Ltd. (72) Inventor Hideo Arai Kodaira, Tokyo 5-20-1, Josuihoncho Hitachi, Ltd. System LSI Development Center Co., Ltd. (72) Inventor Takashi Nishimura 5-2-1, Josuihoncho, Kodaira-shi, Tokyo System LSI Development Center, Hitachi, Ltd. Of which F-term (reference) SS24

Claims (10)

[Claims] 1. A multiplexing method for dividing a plurality of bit streams obtained by encoding a moving image and an accompanying audio signal into packets, and time-division multiplexing the plurality of bit streams in packet units. In regard to the coded bit stream, a bit stream obtained by encoding the moving image is divided so that a specific code indicating the start of coded data of each frame of the moving image to be searched in the decoding process is arranged at the head of the packet. A multiplexing method. 2. A multiplexing method for dividing a plurality of bit streams obtained by encoding a moving picture and an accompanying audio signal into packets, and time-division multiplexing the plurality of bit streams in packet units, wherein at least the moving picture is encoded. The moving image so that a specific code indicating the start of the coded data of each frame of the moving image to be searched for in the decoding process is contained in one packet without dividing the packet into a plurality of packets. A multiplexing method characterized by dividing a bit stream obtained by encoding an image. 3. A multiplexing apparatus which divides a plurality of bit streams obtained by coding a moving image and an accompanying audio signal into packets, and time-division multiplexes the plurality of bit streams in packet units. A start code detecting means for monitoring the bit stream obtained, and detecting a start code indicating the start of encoded data of each frame of a moving image to be searched at the time of decoding processing; A stream is divided by a fixed code length, and when a start code is detected, a bit stream dividing means for dividing a video stream immediately before the start code is provided even if the start code is shorter than the fixed code length. A bit stream obtained by encoding the moving image so as to be arranged at the head of A multiplexing device, characterized in that: 4. A multiplexing method for dividing a plurality of bit streams obtained by encoding a moving image and accompanying audio signals into packets, and time-division multiplexing the plurality of bit streams in packet units, wherein at least the moving image is encoded. Monitoring the converted bit stream, a start code detecting means for detecting a start code indicating the start of the encoded data of each frame of the moving image to be searched in the decoding process, and when the start code is not detected, When the bit stream is divided by a fixed code length and a start code is detected, a bit stream dividing means is provided for dividing the video stream immediately before the start code even if the start code is shorter than the fixed code length. The start code indicating the start of the encoded data of each frame of the moving image to be searched is A multiplexing apparatus characterized in that a bit stream obtained by encoding the moving image is divided so as to be included in one packet without being divided into packets. 5. A moving picture coding apparatus for coding a moving picture by using a variable length code, wherein a specific code indicating the start of the coded data of each frame of the moving picture to be searched for in the decoding process is represented by a bit. So that it is located at an integer multiple of a predetermined number of bits from the beginning of the stream,
A moving picture coding apparatus comprising means for inserting a code word to be discarded in decoding processing immediately before said specific code. 6. The moving picture coding apparatus according to claim 5, wherein said predetermined number of bits is made to match a signal width of a bit stream signal output by said moving picture coding apparatus. 7. A moving picture coding apparatus for coding a moving picture, wherein a signal width matching a code length of a specific code indicating a start of coded data of each frame of the moving picture to be searched for in a decoding process. A moving image encoding apparatus, comprising: means for outputting a bit stream signal. 8. A moving picture coding apparatus comprising a coding means for coding a moving picture and a transmission buffer for temporarily storing coded data, the number of times of outputting coded data until a specific code is output. A moving picture coding apparatus comprising means for generating information. 9. A moving picture coding apparatus comprising: coding means for coding a moving picture; and a transmission buffer for temporarily storing coded data, wherein the transmission buffer is divided into a plurality of areas, and a specific code A moving image encoding apparatus comprising means for outputting a signal indicating presence / absence of each divided area. 10. A recording / reproducing apparatus comprising an image encoding / decoding apparatus, an audio encoding / decoding apparatus, a multiplexing / demultiplexing apparatus, and recording / reproducing means, wherein the image encoding / decoding apparatus is 10. A recording / reproducing apparatus comprising: the video encoding apparatus according to claim 5; and the multiplexing / demultiplexing apparatus including the multiplexing apparatus according to claim 3 or 4. apparatus.