JP3177825B2 - Media coding device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] The present invention relates to speech, text,
The present invention relates to a media encoding device that efficiently encodes and multiplexes various media such as still images and moving images. In the case where various input media signals are encoded with high efficiency and transmitted as a constant-rate bit stream, the encoded output bit stream is temporarily stored in a buffer, read out from the buffer at a constant rate, and transmitted. In this case, for example, the quantization step is switched so that the buffer does not underflow or overflow. When the amount of generated information becomes extremely small, an underflow occurs even by switching the quantization step. In this case, dummy data is inserted. In addition, a method of multiplexing a plurality of input media signals with high efficiency encoding is known, and it is demanded to perform such multiplexing processing efficiently.

[0002]

2. Description of the Related Art FIG. 6 is an explanatory diagram of a conventional media encoding apparatus, in which various input media signals IN1 to INn are highly efficient encoded by media encoding sections 61-1 to 61-n, respectively. As described above, when the amount of information generated by the encoding process is less than a predetermined amount of information, dummy data is inserted and bit streams BS1 to BS
n is input to the multiplexing unit 62, multiplexed according to a multiplexing unit such as a packet unit or a frame unit, and transmitted as a bit stream at a constant rate.

When an input media signal is, for example, a moving image signal, a media encoding unit, for example, uses an MPEG (Moving) signal.
Picture Experts Group) encoder can be applied. In encoding such a moving image signal,
When the screen is flat and has no motion, the amount of information generated by the encoding process is extremely small.

Therefore, in such a case, a method of inserting dummy data as stuffing bits and stuffing bytes is adopted. Also, the stuff code of the macroblock address MBA can be inserted as dummy data. Such dummy data is inserted in each of the media encoding units 61-1 to 61-n, and is input to the multiplexing unit 62 as bit streams BS1 to BSn at a constant rate.

[0005]

In the conventional media encoding units 61-1 to 61-n, it is determined whether or not the amount of information generated by the encoding process is less than a predetermined amount.
When the amount of information is less than a predetermined amount, dummy data is inserted. As a result, the multiplexing section 62 inputs the bit stream at a constant rate and outputs the bit stream at a constant rate even after multiplexing. In this case, for example, in the encoding process of a moving image, it is possible to determine whether or not the amount of generated information is less than a predetermined information amount after the encoding process of one frame. In this case, dummy frame insertion processing is performed by temporarily suspending the encoding processing of the frame.

[0006] Since dummy data is inserted by temporarily interrupting the encoding process, if the number of inserted dummy data is increased, the encoding process is delayed correspondingly. In the case of a signal, there is a problem in that the reproduced image quality is deteriorated due to dropped frames, and in the case of an audio input media signal, there is a problem in that a reproduced sound is interrupted. It is an object of the present invention to efficiently multiplex a plurality of media encoded bit streams.

[0007]

The media encoding apparatus of the present invention (1) encodes input media signals IN1 to INn and outputs bit streams B1 to Bn, and the amount of information generated by the encoding process is reduced. Media encoding units 1-1 to 1-n that output dummy data insertion signals DS1 to DSn when the amount of information is less than a predetermined amount, and bit streams B1 from a plurality of media encoding units 1-1 to 1-n. ~ Bn
And the dummy data insertion signals DS1 to D
A multiplexing unit 2 for inserting and multiplexing dummy data generated according to Sn and outputting a bit stream at a constant rate. Therefore, since dummy data is inserted in the multiplexing unit 2, a plurality of media encoding units 1-
1 to 1-n, it is possible to achieve economy compared with the configuration in which dummy data is inserted, and the encoding processing in each of the media encoding units 1-1 to 1-n is interrupted. You don't have to.

(2) Input media signals IN1 to INn
And outputs bit streams B1 to Bn and outputs dummy data insertion signals DS1 to DSn and insertion number data when the amount of information generated by the encoding process is less than a predetermined amount of information. Parts 1-1 to 1-
n and bit streams B1 to Bn from a plurality of media encoding units 1-1 to 1-n, and multiplex by inserting dummy data of a number generated according to a dummy data insertion signal and insertion number data. And a multiplexing unit 2 for outputting a constant rate bit stream.

(3) Input media signals IN1 to INn
And outputs bit streams B1 to Bn, and outputs at least data on the number of insertions of dummy data when information generated by the encoding processing is less than predetermined information. n, a buffer memory for temporarily storing insertion number data, a detecting unit for detecting a specific position in a bit stream from a plurality of media encoding units, and a plurality of media encoding units 1-1 to 1-1.
-N, and multiplexes the bit streams B1 to Bn from the n. A multiplexing unit 2 for multiplexing and outputting a bit stream at a constant rate.

[0010] (4) The multiplexing unit 2 includes a packetizing unit that converts the bit streams from the plurality of media encoding units into packets of a fixed length, and a dummy data generating unit according to a dummy data insertion signal or insertion number data. And a dummy data insertion control unit for controlling the insertion timing of the dummy data.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a diagram for explaining the principle of the present invention.
The media encoding units 1-1 to 1-n corresponding to Nn perform bit stream B1 to B
n, and a dummy data determination unit 4 that determines whether the amount of generated information is less than a predetermined amount of information, and outputs the dummy data insertion signals DS1 to DSn when the amount of generated information is less than the predetermined amount of information. Have.

The multiplexing unit 2 outputs a dummy data insertion signal D
A dummy data processing unit 6 for generating dummy data according to S1 to DSn;
A multiplexing unit 5 for multiplexing Bn and inserting and multiplexing dummy data from the dummy data processing unit 6 is provided.

The input media signals IN1 to INn are encoded in an encoding processing unit 3 of the media encoding units 1-1 to 1-n, and whether or not the amount of information generated at that time is less than a predetermined information amount. This is determined by the dummy data determination unit 4, and if not, dummy data insertion signals DS1 to DSn are output.

For example, assuming that the media encoding unit is an MPEG encoding unit, the predetermined information amount is Do bytes / frame, and the number of dummy data inserted by one pulse of the dummy data insertion signals DS1 to DSn is nd bytes. When the encoding of each frame is completed, the actual generated information amount Df bytes of one frame is obtained. If Df ≧ Do, no dummy data insertion signal is generated, but if Df <Do, the dummy data insertion signal is not generated. Are output as (Do−Df) / nd pulses. Dummy data processing unit 6 of multiplexing unit 2
Generates dummy data of the number of bytes corresponding to the number of pulses of the dummy data insertion signal, and multiplexes the generated dummy data with the multiplexing processing unit 5 to output a bit stream of a constant rate.

As described above, the media encoding units 1-1 to 1-1
1-n does not perform the dummy data insertion processing, so that the encoding processing unit 3 can continuously perform the encoding processing on the input media signals IN1 to INn. Therefore, the possibility of dropping frames is small, the deterioration of the reproduced image quality can be prevented, and the possibility that the reproduced sound is interrupted can be reduced.

The multiplexing unit 2 includes a media encoding unit 1-1.
The dummy data is generated in the dummy data processing unit 6 in accordance with the dummy data insertion signals DS1 to DSn to 1-n, and the dummy data is added to the multiplexing processing unit 5, and the bit streams B1 to Bn are converted into packet units or frames. Dummy for multiplexing in units of multiplexing and inserting dummy data into a bit stream with a small amount of generated information and outputting as a multiplexed bit stream at a constant rate. By centralizing the data insertion processing, hardware and software can be made more economical.

FIG. 2 is an explanatory diagram of a main part of the first embodiment of the present invention, showing the function of the multiplexing unit 2 of FIG. 1, wherein 11 is a dummy data insertion control unit, and 12 is a dummy data insertion control unit. A data generating unit, 13 is a packetizing unit, 14 is a multiplexing selecting unit constituting a main part of a multiplexing processing unit for multiplexing bit streams, and 15 is a part of a multiplexing processing unit for inserting and multiplexing dummy data. The insertion selection units 16-1 to 16-n are configured as buffer memories (F) corresponding to the bit streams B1 to Bn.
IFO).

The dummy data insertion control unit 1 shown in FIG.
1 and the dummy data generation unit 12 constitute the dummy data processing unit 6 of FIG.
The multiplexing processing unit 5 of FIG. 1 is composed of n, the multiplexing selection unit 14, the insertion selection unit 15, and the packetization unit 13.
The buffer memories 16-1 to 16-n are first-in first-out memories (FIFO) for holding the encoded bit stream until it is selected by the multiplexing selection unit 15.

The dummy data insertion signals DS1 to DSn are
The bit streams B1 to Bn of the coded output added to the dummy data insertion control unit 11 are stored in the buffer memory 16
-1 to 16-n are applied to the multiplexing selection unit 14. The multiplexing selection unit 14 is controlled by a control unit (not shown), and selects according to the storage occupancy of the buffer memories 16-1 to 16-n. The bit stream is read out from .about.16-n, is packetized by the packetizing unit 13, and is normally output as a constant-rate bit stream via the insertion selecting unit 15 in the state shown in the figure.

The dummy data insertion control unit 11 controls the dummy data generation unit 12 according to the dummy data insertion signals DS1 to DSn, and controls the insertion selection unit 15 so that the dummy data generation unit 12 generates the dummy data. The data is inserted and multiplexed to form a constant-rate bit stream.
In this case, the insertion selection unit 15 is omitted, and the dummy data insertion control unit 11 controls the dummy data generation unit 12 and the packetization unit 13 so that the dummy data from the dummy data generation unit 12 is indicated by a dotted arrow. , Packetizer 1
3 to be packetized and stored in the buffer memories 16-1 to 16-1.
16-n and multiplexed by inserting into the bit stream from
It is also possible to output as a constant-rate bit stream.

FIGS. 3A and 3B are explanatory diagrams of dummy data insertion according to the first embodiment of the present invention. FIG. 3A shows an encoded bit stream of the medium A, and FIG. Dummy data insertion signal, (c) is a coded bit stream of medium B, (d) is a dummy data insertion signal obtained by encoding processing of medium B, and (e) is a dummy data inserted and multiplexed output. data,
(F) shows the format of the packet.

For example, the generated information amount Df for one frame of the bit streams A1 and A2 by encoding the medium A
Is smaller than the predetermined information amount Do, and (Do−Df) / nd
= 4 (numbers), four pulses of the dummy data insertion signal are output as shown in (b). Also, one of the bit streams B1 and B2 by encoding one frame of the medium B
Also for the frames, the generated information amount Df is smaller than the predetermined information amount Do, and (Do−Df) / nd = 2 (number)
In this case, two pulses of the dummy data insertion signal are output as shown in FIG.

In FIG. 2, when the dummy data insertion signals DS1 to DSn are input from the media encoding unit, the dummy data insertion control unit 11 immediately controls the dummy data generation unit 12 to FIG. 3 shows a case where dummy data of a number corresponding to the number of pulses of the data insertion signal is generated and inserted when multiplexing bit streams via the buffer memories 16-1 to 16-n, and FIG. FIG. 3E shows four pieces of dummy data D according to four pulses of the dummy data insertion signal of FIG. 3B after multiplexing of bit streams A1 and B1 of media A and B.
Are inserted, and two pieces of dummy data D are inserted according to two pulses of the dummy data insertion signal shown in (d). Thereafter, bit streams A2 and B2 of media A and B are inserted.
Indicates multiplexed output data. When the dummy data insertion signal is 0 pulse, bit streams A2, B
2, A3, B3,... Are multiplexed without inserting dummy data.

The format shown in FIG. 3 (f) is audio,
In the case of dummy data, a type TP indicating a media type such as a moving image or a still image, a number of data L, and data D1 to Dm are shown. To Dm can be all zero. Therefore, when the multiplexed bit stream is demultiplexed on the receiving side and decoded for each media, dummy data can be identified and deleted.

FIG. 4 is an explanatory view of a main part of a second embodiment of the present invention, wherein 20-1 to 20-n are media encoding units,
21 is a dummy data insertion control unit, 22 is a dummy data generation unit, 23 is a packetization unit, 24 is a multiplex selection unit, 25
Is an insertion selection unit, 26-1 to 26-n are buffer memories (FIFO), IN1 to INn are input media signals, B
1 to Bn are bit streams of encoded output, DS1 to D
Sn indicates a dummy data insertion signal, and DN1 to DNn indicate insertion number data.

In this embodiment, the media encoding unit 20
-1 to 20-n, input media signals IN1 to I-I
Nn is encoded, bit streams B1 to Bn are output, and it is determined whether the amount of information generated at that time is less than a predetermined amount of information. If not, dummy data indicating that dummy data is to be inserted Insertion signals DS1 to DSn
And insertion number data DN1 to DNn indicating the number of insertions
And output.

The point that the multiplexing is performed by the multiplexing selection unit 24 by selection control corresponding to the storage occupancy of the buffer memories 26-1 to 26-n is the same as in the first embodiment. Further, the dummy data insertion control unit 21 includes a dummy data generation unit 22 and an insertion selection unit based on the dummy data insertion signals DS1 to DSn from each of the media encoding units 20-1 to 20-n and the insertion number data DN1 to DNn. 25, the dummy data generation unit 22 generates dummy data in accordance with the insertion number data DN1 to DNn, switches the insertion selection unit 25 from the solid line position to the dotted line position, inserts the dummy data, and It is output as a rate bit stream.

The dummy data generated by the dummy data generating unit 22 can be added to the packetizing unit 23 as shown by the dotted arrow, and the dummy data can be packetized and inserted. In that case, the function of the insertion selection unit 25 is provided in the multiplex selection unit 24 or the packetization unit 23.

FIG. 5 is an explanatory view of a main part of a third embodiment of the present invention, wherein 31 is a dummy data insertion control section, 32 is a dummy data generation section, 33 is a packetization section, and 34 is a multiplexing selection section. , 35 are insertion selecting units, and 36-1 to 36-n are buffer memories (FIs) for adding bit streams B1 to Bn.
FO) and 37-1 to 37-n are buffer memories (FIFOs) for adding dummy data insertion number data DN1 to DNn.
D), 38-1 to 38-n are detection units for detecting a dummy data insertion position, IN1 to INn are input media signals, D
T1 to DTn indicate detection signals.

The media encoding units 30-1 to 30-n
Bit streams B1 to Bn obtained by encoding the input media signals IN1 to INn are stored in buffer memories 36-1 to 36-36.
n, it is determined whether or not the amount of generated information at that time satisfies a predetermined amount of information. If not, the number of inserted dummy data is determined, and the number of inserted data DN1
To DNn are added to the buffer memories 37-1 to 37-n.

The multiplexing selection section 34 sequentially selects the bit stream according to the storage occupancy of the buffer memories 36-1 to 36-n and the like, as in the above-described embodiments, and converts the bit stream into packetization sections 33. In addition, the bit stream is packetized and multiplexed. In each of the embodiments described above, the dummy data is inserted between arbitrary packets of the bit stream by the dummy data insertion signal. In this embodiment, the dummy data is inserted at a specific position by the bit stream. For that purpose, a specific position where dummy data is to be inserted is detected by the detectors 38-1 to 38-n, and the detection signals DT1 to DTn are added to the dummy data insertion controller 31.

For example, an input media signal is a moving image signal, and an I picture (Intra-Pic) in MPEG encoding is used.
), P picture (Predictive-Picture), B picture (Bidirectionally Predictive-Picture)
When inserting dummy data in the case where the amount of information generated in the previous encoding is less than the predetermined amount of information at the beginning of a picture such as "00000", the detection units 38-1 to 38-n use "00000".
A picture start code (PSC) having a 32-bit configuration by repeating 100 "is detected, and its detection signal DT1 is detected.
To DTn to the dummy data insertion control unit 31.

The dummy data insertion control unit 31 controls the buffer memories 37-1 to 37-3 according to the detection signals DT1 to DTn.
7-n, the dummy data insertion number data DN1 to DNn are read. The insertion number data D
When N1 to DNn are 0, that is, when the generated information amount satisfies the predetermined information amount, the dummy data insertion control unit 31 does nothing.

If the number-of-insertions data DN1 to DNn is equal to or greater than 1, the dummy-data insertion control unit 31 causes the dummy-data generation unit 32 to generate dummy data according to the number-of-insertions data. Is switched to the position indicated by the dotted line, and when the dummy data from the dummy data generation unit 32 is inserted, the insertion selection unit 35 is switched to the position indicated by the solid line. Thereafter, since the bit stream including the picture start code (PSC) is packetized and transmitted by the packetizing unit 33, dummy data for the previous picture is inserted at the head of the picture. Thus, the decoding unit inserts the dummy data between the pictures, so that the decoding process for the picture can be continued without being interrupted by the dummy data. Therefore, the processing of the decoding unit becomes easy.

The present invention is not limited to the above-described embodiments, but can be variously added and changed. For example, in each embodiment, the dummy data Although insertion is performed collectively according to the insertion signal or insertion number data, it is also possible to insert in a multiplexed bit stream in a distributed manner, for example, between packets. The specific position where the dummy data is inserted, such as the head position of the picture, can be selected according to the media type. For example, a frame synchronization signal in the case of audio is detected, and the dummy data is inserted before that. It can also be a specific position.

[0036]

As described above, according to the present invention, in the media encoding units 1-1 to 1-n for encoding various input media signals IN1 to INn, the amount of generated information is a predetermined amount of information. In this case, the dummy data is not inserted into the multiplexing unit 2 and the dummy data insertion signal is added to the multiplexing unit 2 so that the multiplexing unit 2 inserts and multiplexes the dummy data. 1 to 1-n, the encoding process can be continued and a sufficient amount of dummy data can be inserted in the multiplexing unit 2, so that a constant rate bit stream can be output stably. There are advantages that can be. Further, since dropping of frames in the case of a moving image does not occur, there is an advantage that deterioration in reproduction image quality is avoided and reproduction sound is not interrupted. Also, dummy data can be inserted at a specific position, thereby facilitating the decoding process.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is an explanatory view of a main part of the first embodiment of the present invention.

FIG. 3 is an explanatory diagram of dummy data insertion according to the first embodiment of this invention.

FIG. 4 is an explanatory view of a main part of a second embodiment of the present invention.

FIG. 5 is an explanatory view of a main part of a third embodiment of the present invention.

FIG. 6 is an explanatory diagram of a conventional media encoding device.

[Explanation of symbols]

 1-1 to 1-n Media encoding unit 2 Multiplexing unit 3 Encoding processing unit 4 Dummy data determining unit 5 Multiplexing processing unit 6 Dummy data processing unit IN1 to INn Input media signal B1 to Bn Bit stream DS1 to DSn Dummy Data insertion signal

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int. Cl. ⁷ , DB name) H04J 3/00-3/26 H04N 7/24

Claims (4)

(57) [Claims] A media encoding unit that encodes an input media signal to output a bit stream, and outputs a dummy data insertion signal when an amount of information generated by the encoding process is less than a predetermined amount of information; A multiplexing unit that multiplexes the bit streams from the plurality of media encoding units, inserts and multiplexes the dummy data generated according to the dummy data insertion signal, and outputs a bit stream at a constant rate. A media encoding device characterized by the above-mentioned. 2. A medium that encodes an input media signal and outputs a bit stream, and outputs a dummy data insertion signal and insertion number data when the amount of information generated by the encoding process is less than a predetermined amount of information. An encoding unit, which multiplexes the bit streams from the plurality of media encoding units, and inserts and multiplexes the number of dummy data generated according to the dummy data insertion signal and the insertion number data, and And a multiplexing unit that outputs the bit stream of the media encoding. 3. A media encoding unit for encoding an input media signal to output a bit stream, and outputting at least data on the number of inserted dummy data when information generated by the encoding process is less than predetermined information. A buffer memory for temporarily storing the insertion number data, a detection unit for detecting a specific position of the bit stream from the plurality of media encoding units, and a bit stream from the plurality of media encoding units. Along with multiplexing, the number of dummy data generated by reading the insertion number data from the buffer memory based on the detection signal that has detected the specific position by the detection unit is inserted into the specific position and multiplexed, and a constant rate A media encoding device comprising: a multiplexing unit that outputs a bit stream. 4. The multiplexing unit includes: a packetizing unit configured to convert the bit streams from the plurality of media encoding units into packets of a fixed length; a dummy data generating unit according to the dummy data insertion signal or the insertion number data; 4. The media encoding apparatus according to claim 1, further comprising: a dummy data insertion control unit that controls a dummy data insertion timing.