JP3322671B2 - MPEG video decoder - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an MPEG (Moving
Picture Expert Group) relates to a video decoder.

[0002]

2. Description of the Related Art CDs (Compact Disks) are widely used all over the world as media for providing high-quality music information.
2. Description of the Related Art In recent years, multimedia applications centered on image information and audio information as well as music information have been promoted. Various CDs used in multimedia are collectively called a CD family. The CD family includes a CD-DA (CD-Digital Audio) family, a so-called music CD, and a CD-RO, a so-called data CD.
M (CD-Read Only Memory) family. C
The D-ROM family includes CD-IFMV (CD-Interac
tive Full MotionVideo) or CD-IDV (CD-Inte
ractive Digital Video). CD-IFMV
Includes video CDs and karaoke CDs.

[0003] The information handled in multimedia is enormous and diverse, and it is necessary to process such information at high speed in order to put the multimedia into practical use. In order to process information at high speed, data compression / decompression technology is indispensable. Compression of such data
MPEG (Moving Picture ExpertGro
up)) method. This MPEG system is based on IS
O (International Organization for Standardizatio
n) / IEC (International ElectrotechnicalCommiss
ion) MPEG committee under ISO / IEC JTC1 / SC29 / WG1
It is being standardized by 1). MPEG is composed of three parts. In Part 1, "MPEG System Part" (ISO / IEC IS 11172 Part1: Systems),
A multiplex structure (multiplex structure) of video data and audio data and a synchronization method are defined. Part 2 “MPEG Video Part” (ISO / IEC IS
11172Part2: Video) specifies a high-efficiency encoding method for video data and a format of the video data. Part 3 “MPEG Audio Part” (ISO / IE
CIS 11172 Part3: Audio) specifies a high-efficiency encoding method for audio data and a format of the audio data.

[0004] By utilizing the MPEG system, CD
-Movie playback is also possible in the ROM family.
Karaoke CD can be converted from CD-IFMV format to MP
Only the portion related to the EG method is extracted, and only the moving image reproduction is performed. Video CD, in addition to video playback,
A menu playback function called PBC (Play Back Control), which enables still image playback and expression combining still image playback and moving image playback, is added. Therefore,
Video CDs are compatible with CD-IFMV and CD-I
A disc in a video CD format can be reproduced by an FMV player.

[0005] It should be noted that at present, MPEG-1 and MPEG-
There are two methods. MPEG-1 is mainly CD-RO
Compatible with storage media such as M family, MP
EG-2 supports a wide range of applications, including MPEG-1. The video data handled by the MPEG video part relates to a moving image, and the moving image is composed of several tens (eg, 30) frames (screens) per second. Video data includes a sequence (Sequence), a GOP (Group Of Pictures), a picture, a slice (Slice), and a macro block (Macrobloc).
k), and has a hierarchical structure of six layers in the order of blocks. MPEG
At -1, a frame corresponds to a picture. MP
In EG-2, a frame or a field can correspond to a picture. Two fields constitute one frame. A structure in which a frame corresponds to a picture is called a frame structure, and a structure in which a field corresponds to a picture is called a field structure.

[0006] MPEG uses a compression technique called inter-frame prediction. Inter-frame prediction compresses data between frames based on temporal correlation. In the inter-frame prediction, bidirectional prediction is performed. The bidirectional prediction is to use both forward prediction for predicting a current reproduced image from a past reproduced image (picture) and backward prediction for predicting a current reproduced image from a future reproduced image.

[0007] This bidirectional prediction is based on an I-picture (Intra-Pi
), a P picture (Predictive-Picture), and a B picture (Bidirectionally predictive-Picture). The I picture is generated independently of a past or future reproduced image. The P picture is generated by forward prediction (prediction from a past I picture or P picture). B pictures are generated by bidirectional prediction. In bidirectional prediction, a B picture is generated by any one of the following three predictions. Prediction from past I or P pictures, prediction from future I or P pictures, past and future I or P pictures
Prediction from pictures. Then, these I, P, and B pictures are respectively encoded. That is, an I picture is generated even if there is no past or future picture. On the other hand, a P picture is not generated without a past picture,
A B picture is not generated without a past or future picture.

In the inter-frame prediction, first, an I picture is periodically generated. Next, a frame several frames ahead of the I picture is generated as a P picture. This P
The picture is generated by one-way (forward) prediction from the past to the present. Subsequently, a frame located before the I picture and after the P picture is generated as a B picture. When generating this B picture, an optimal prediction method is selected from three of forward prediction, backward prediction, and bidirectional prediction. In general, in a continuous moving image, a current image and images before and after the current image are very similar, and only a part thereof is different. Then, the previous frame (for example, I
Assuming that the picture and the next frame (for example, a P picture) are the same, if there is a change between both frames, only the difference (B picture data) is extracted and compressed.
Thus, data between frames can be compressed based on temporal correlation.

The data sequence (bit stream) of video data encoded in accordance with the MPEG video part is called an MPEG video stream (hereinafter, abbreviated as video stream). FIG. 5 shows a conventional MP.
2 shows a main block circuit of the EG video decoder 101.
The MPEG video decoder 101 includes a bit buffer 10
2. Picture header detection circuit 103, MPEG video decode core circuit (hereinafter abbreviated as decode core circuit) 10
4, an overflow detection circuit 105 and a picture skip circuit 106.

[0010] An MPEG video stream (hereinafter abbreviated as a video stream) transferred from a transmission medium (not shown) is input to a bit buffer 102. The bit buffer 102 is configured by a ring buffer including a random access memory (RAM) having a FIFO (First-In-First-Out) configuration, and sequentially accumulates video streams.

The transmission medium includes a LAN (Local Ar
ea Network), communication media such as video CD and DVD
(Digital Video Disk) and VTR (Video Tape Rec)
order) and broadcast media such as terrestrial broadcasting, satellite broadcasting, and CATV (Community Antenna Television). Picture header detection circuit 10
No. 3 detects a picture header from the video stream stored in the bit buffer 102. On the basis of the detection result of the picture header detection circuit 103, a video stream for one picture is read from the bit buffer 102 at regular intervals.

Each picture read from the bit buffer 102 is transferred to the decode core circuit 104 via the picture skip circuit 106. The picture skip circuit 106 has two nodes 106a and 106b. Then, the picture skipping circuit 106
06a, each picture read from the bit buffer 102 is directly decoded by the decode core circuit 10a.
Transfer to 4. When connected to the node 106b, each picture read from the bit buffer 102 is skipped in units of pictures without being transferred to the decode core circuit 104. The switching operation of each of the nodes 106a and 106b of the picture skip circuit 106 is performed by the picture header detection circuit 103 and the overflow detection circuit 10
5.

The decode core circuit 104 converts each picture transferred from the picture skip circuit 106 into an MPE.
By decoding in accordance with the G video part, a video output for each picture is generated, and the video output is output to an externally provided display (not shown). And
The display reproduces a moving image based on the video output.

The overflow detection circuit 105 detects whether the bit buffer 102 has overflowed. Then, the overflow detection circuit 105
When it detects that the bit buffer 102 has overflowed, it controls the picture skip circuit 106 to switch the connection to the node 106b side, and each picture read from the bit buffer 102 until the bit buffer 102 no longer overflows. Let me skip.
Thereafter, upon detecting that the overflow of the bit buffer 102 has been released, the overflow detection circuit 105 controls the picture skip circuit 106 to
The connection is switched to the 06a side, and each picture read from the bit buffer 102 is transferred to the decode core circuit 104.

Here, the bit rate RB of the video stream transferred from the transmission medium is fixed. Therefore, it is necessary to control the occupancy of the bit buffer 102 so that the amount of data for one picture is not too large or too small, and the bit buffer 102 does not overflow or underflow. Therefore,
In the MPEG video part, a virtual MPEG video decoder is assumed, and a definition for it is made.

FIG. 6 shows a change in the occupancy of the bit buffer 102 during normal reproduction. Bit buffer 10
The occupancy Bm of 2 rises with the bit rate RB as the slope of the graph. The bit rate RB is defined as shown in equation (1) according to the BR (Bit Rate) of the sequence header at the beginning of the sequence. Also, the picture rate R of the video stream transferred from the transmission medium
P is defined by the PR (Picture Rate) of the sequence header. Then, the capacity B of the bit buffer 102
Is the VBV (Vbv [Video Buffering
Verifier] and BufferSize) as shown in equation (2). Then, for each frame period, a video stream for one picture that the decode core circuit 104 is to decode at that time is read from the bit buffer 102 at a stretch. Here, the data amount X of the video stream transferred from the transmission medium and input to the bit buffer 102 during one frame period is equal to the bit rate RB
And the picture rate RP is defined as shown in equation (3). Therefore, the occupation amount Bm (= B0 to B6) of the bit buffer 102 immediately after the video stream for one picture is read from the bit buffer 102 at a stretch.
Is defined based on the data amount X and the capacity B of the bit buffer 102 so as to satisfy the condition shown in Expression (4).

RB = 400 × BR (1) B = 16 × 1024 × VBV (2) X = RB / RP (3) 0 <Bm <BX = B− (RB) / RP) (4) The bit buffer 10 is set so as to satisfy the condition shown in the equation (4).
If the occupation amount Bm of 2 is specified, the bit buffer 1
02 does not overflow or underflow. Conversely, when the occupation amount Bm of the bit buffer 102 exceeds the threshold value (BX), the possibility that the bit buffer 102 overflows due to the video stream input to the bit buffer 102 in the next one frame period is extremely high. Become.

At the time of normal reproduction, the bit rate RB and the picture rate RP are set so that the equation (4) is satisfied.
, And the capacity B are defined. Therefore, equation (2)
If the capacity B of the bit buffer 102 is set as shown in (1), the bit buffer 102 should not overflow or underflow. However, as described above, the occupancy of the bit buffer 102 (Bm
/ B), the bit buffer 102 may overflow. Overflow detection circuit 1
05 and the picture skip circuit 106 are provided for this reason.

That is, the overflow detection circuit 1 determines whether the bit buffer 102 overflows.
If the overflow is detected in step S05 and the overflow occurs, each picture read from the bit buffer 102 is skipped via the picture skip circuit 106. As a result, the overflow of the bit buffer 102 is released.

[0020]

SUMMARY OF THE INVENTION Bit buffer 102
Is overflowed with a ring buffer. Therefore, when an overflow occurs, the video stream already stored in the bit buffer 102 is overwritten by the newly input video stream. as a result,
The video stream already stored in the bit buffer 102 is destroyed and lost.

For example, when the bit buffer 102 overflows during the decoding of an arbitrary picture in the decode core circuit 104, a portion of the picture being decoded that is left in the bit buffer 102 is newly input. The video stream is overwritten. As a result, the portion of the picture being decoded that remains in the bit buffer 102 is destroyed and lost. Then, the decoding core circuit 104 cannot complete the decoding of the picture, and cannot generate a video output of the picture.

As described above, a P picture cannot be generated without a past picture, and a B picture cannot be generated without a past or future picture. Only I pictures can be generated without past or future pictures. Therefore, if the picture being decoded at the time when the bit buffer 102 overflows is an I picture or a P picture, of the pictures in the video stream stored in the bit buffer 102, Cannot decode all P pictures and B pictures up to the I picture read out. That is, in the decode core circuit 104, the bit buffer 1
The decoding process cannot be restarted from 02 until the next I picture is read.

As described above, when the bit buffer 102 overflows, a large number of pictures cannot be decoded, and thus frames are dropped in a moving image reproduced for the pictures. As a result, the motion of the moving image is not smooth but jerky, and the image quality is deteriorated to make it hard to see. By the way, the picture skip circuit 106 skips the picture read from the bit buffer 102 regardless of the type of the picture. Therefore, the picture skipped from the picture skip circuit 106 may be an I picture or a P picture. In this case, when the overflow of the bit buffer 102 is released and the decoding operation in the decoding core circuit 104 is restarted, the decoding process cannot be performed on several pictures following the skipped picture, and the video output is generated. Will stop for a while. As a result, frames dropped in the video played on the display,
The motion of the video becomes jerky, the image quality deteriorates and it becomes hard to see.

As described above, the bit buffer 10
Even if the occupation ratio of 2 is controlled, the bit buffer 102 may underflow. For example, when a video CD is used as a transmission medium, a video stream recorded on a disc may not be able to be read due to scratches or vibration of the disc. In that case,
Until the new video stream is transferred from the video CD and the underflow of the bit buffer 102 is released, the decoding operation in the decoding core circuit 104 must be stopped, and the generation of the video output is also stopped. . As a result, frames are dropped in the moving image reproduced on the display, the motion of the moving image becomes jerky, the image quality is deteriorated, and it becomes hard to see.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and it is an object of the present invention to provide an MPEG video decoder which can reduce the number of dropped frames in a reproduced moving image and improve the visibility. For that purpose.

[0026]

[0027]

[0028]

An MPEG video decoder according to claim 1 stores a video stream transferred from an external device, and reads out the video stream for one picture at a time, and reads the video stream from the bit buffer. A picture header detection circuit for determining the type of the picture based on the picture header of the video stream obtained, an MPEG video decoding core circuit for decoding each picture in accordance with the MPEG video part,
An overflow determination circuit for detecting an occupancy of the bit buffer and comparing the occupancy with a threshold; and an I picture or a P picture being read from the bit buffer when the occupancy of the bit buffer exceeds a threshold. The gist of the invention is to provide a picture skip circuit for transferring the picture as it is to the MPEG video decoding core circuit and skipping the picture when the B picture is read without transferring the picture to the MPEG video decoding core circuit.

The MPEG video decoder according to the second aspect sequentially accumulates video streams transferred from the outside,
A bit buffer in which the video stream is read out by one picture at regular intervals; and a picture header detected from the video stream stored in the bit buffer, and a picture read out from the bit buffer based on the picture header. A picture header detection circuit that determines the type of the video buffer, an MPEG video decoding core circuit that generates a video output for each picture by decoding each picture in accordance with the MPEG video part, and detects the occupancy of the bit buffer. An overflow judging circuit for comparing the occupation amount and a threshold value, and a first and second node. When connected to the first node side, each picture read from the bit buffer is directly used as the MPEG video. Transfer to the decode core circuit and connect to the second node A picture skip circuit that skips each picture read from the bit buffer in units of pictures without transferring the picture to the MPEG video decode core circuit. In this case, when an I picture or a P picture is read from the bit buffer, the connection is made to the first node side, and when a B picture is read, the connection is made to the second node side. .

According to a third aspect of the present invention, in the MPEG video decoder according to the first or second aspect , the capacity (BA) of the bit buffer is a buffer size (Vbv Buffer S).
ize) and a value (X) obtained by dividing a bit rate (picture rate) by a picture rate (picture rate) to a capacity (B) specified in accordance with the buffer size, the bit rate, and the picture rate. The gist is that each is specified by a sequence header attached to the head of the video stream sequence.

According to a fourth aspect of the present invention, in the MPEG video decoder according to any one of the first to third aspects, the capacity (BA) of the bit buffer is a buffer size (Vb).
v Buffer size), set the bit rate (bit rate) to the picture rate (picture
te) is set to exceed the value obtained by adding the value (X) divided by
The gist is that the buffer size, the bit rate, and the picture rate are specified by a sequence header attached to the head of the sequence of the video stream.

A fifth aspect of the MPEG video decoder according to the third or fourth aspect is that the threshold is set to the same value as the capacity (B).

[0033]

(First Embodiment) A first embodiment of the present invention will be described below with reference to FIGS. FIG. 1 shows an MPEG video decoder 1 according to the present embodiment.
2 shows a main part block circuit. MPEG video decoder 1
Is a bit buffer 2, a picture header detection circuit 3, and M
It comprises a PEG video decode core circuit (hereinafter abbreviated as a decode core circuit) 4, an overflow determination circuit 5, and a picture skip circuit 6.

An MPEG video stream (hereinafter abbreviated as video stream) transferred from a transmission medium (not shown) is input to a bit buffer 2. The transmission media include communication media such as a LAN (Local Area Network), and video CDs and DVDs (Digital Video Dis
k) and storage media such as VTR (Video Tape Recorder), terrestrial broadcasting and satellite broadcasting, and CATV (Communication
broadcasting media, such as ity Antenna Television).

The bit buffer 2 is a FIFO RAM
, And sequentially accumulates video streams. Here, the capacity BA of the bit buffer 2 is set to a value obtained by adding the data amount X defined by the equation (3) and an appropriate margin ΔB to the capacity B defined by the equation (2). (BA = B + X + ΔB). For example, in a video CD, the capacity B is 46 kbytes and the data amount X is 6
It is specified in k bytes. The margin ΔB is the capacity B
And about 2 k bytes based on the data amount X. Therefore, the capacity BA of the bit buffer 2 is about 5
It is set to 4k (= 46k + 6k + 2k) bytes.

The picture header detection circuit 3 detects a picture header from the video stream stored in the bit buffer 2. Based on the detection result of the picture header detection circuit 3, a video stream for one picture is read out from the bit buffer 2 at regular intervals. Further, the picture header detection circuit 3 determines the type of the picture read from the bit buffer 2 based on the picture header.

Each picture read from the bit buffer 2 is transferred to the decode core circuit 4 via the picture skip circuit 6. The picture skip circuit 6 has two nodes 6a and 6b. Then, when connected to the node 6a side, the picture skip circuit 6 transfers each picture read from the bit buffer 2 to the decode core circuit 4 as it is. Further, when connected to the node 6b side, each picture read from the bit buffer 2 is skipped in picture units without being transferred to the decode core circuit 4. Each node 6 of the picture skip circuit 6
a, 6b is switched by the picture header detection circuit 3
And an overflow determination circuit 5.

The decode core circuit 4 generates a video output for each picture by decoding each picture transferred from the picture skip circuit 6 in accordance with the MPEG video part, and provides the video output externally. Output to a display (not shown). Then, a moving image based on the video output is reproduced on the display.

The overflow determination circuit 5 detects the occupancy Bm of the bit buffer 2, and compares the occupancy Bm with the first threshold value BTH1 and the second threshold value BTH2. Here, the first threshold value BTH1 is set to the same value as the capacity B (BTH1 = B). The second threshold value BTH2 is set to a value obtained by adding the data amount X to the capacity B (BTH2 = B +
X). That is, the capacity BA of the bit buffer 2 is a value obtained by adding a margin ΔB to the second threshold value BTH2.

Next, the switching operation of the nodes 6a and 6b of the picture skip circuit 6 will be described with reference to the flowchart shown in FIG. First, in step (hereinafter referred to as S) 1, if the overflow determination circuit 5 determines that the occupation amount Bm exceeds the first threshold BTH1 (Bm> BTH1), the process proceeds to S2 and exceeds. If it is determined that there is no (Bm ≦ BTH1), the process proceeds to S3.

In S2, the overflow determination circuit 5
Accordingly, when it is determined that the occupation amount Bm exceeds the second threshold value BTH2 (Bm> BTH2), the processing shifts to S5, and when it is determined that the occupation amount Bm does not exceed the second threshold value BTH2 (Bm ≦ BTH2), S4
Move to. In S4, the picture header detection circuit 3
Determines the type of the picture read from the bit buffer 2, and if the picture is a B picture, S
5. If the picture is an I picture or a P picture, go to S6.
Move to.

In S5, the picture skip circuit 6 is switched to the node 6b, and the picture read from the bit buffer 2 is skipped. Then, the process returns to S1. In S3, the picture skip circuit 6 is switched to the node 6a side, and the picture read from the bit buffer 2 is transferred to the decode core circuit 4. Then, the process returns to S1.

As described above, according to the present embodiment, the following operations and effects can be obtained. If the occupancy Bm of the bit buffer 2 does not exceed the first threshold value BTH1, the picture read from the bit buffer 2 is transferred to the decode core circuit 4 regardless of the type. When the occupation amount Bm is between the first threshold value BTH1 and the second threshold value BTH2, when an I picture or a P picture is read from the bit buffer 2, the picture is transferred to the decode core circuit 4, and the B picture is Is read, the picture is skipped. When the occupation amount Bm exceeds the second threshold value BTH2, the picture read from the bit buffer 2 is skipped regardless of the type.

As described above, the occupation amount Bm is equal to the first threshold value B.
If it is between TH1 and the second threshold BTH2, the B picture is skipped preferentially among the pictures read from the bit buffer 2. As a result, the occupation amount Bm of the bit buffer 2 decreases, and the overflow hardly occurs. Here, as described above, a B picture is generated by bidirectional prediction, and its importance is lower than that of an I picture or a P picture.

Therefore, even if the B picture read from the bit buffer 2 is skipped,
The decoding process can be performed in the decoding core circuit 4 on the picture read from the decoder irrespective of its type. As described above, when the occupation amount Bm exceeds the second threshold value BTH2, the picture read from the bit buffer 2 is skipped regardless of the type. As a result, the occupation amount Bm of the bit buffer 2 is reduced and overflow does not occur.

A margin Δ is added to the capacity BA of the bit buffer 2.
Since B is provided, overflow of the bit buffer 2 is more unlikely to occur. Here, as the margin ΔB is larger, the overflow of the bit buffer 2 is less likely to occur, but the capacity BA of the bit buffer 2 becomes larger, so that the cost increases and the economic efficiency deteriorates. Therefore, it is necessary to determine the optimum margin ΔB by performing experiments for actually processing various video streams.

In this embodiment, before the bit buffer 2 overflows, the picture read from the bit buffer 2 is skipped based on its type and the occupation amount Bm to prevent the occurrence of overflow. Has been prevented. On the other hand, in the conventional example, after the overflow occurs in the bit buffer 2, the overflow is canceled by skipping the picture read from the bit buffer 2 regardless of its type. Therefore, according to the present embodiment, the number of dropped frames occurring in a moving image reproduced on the display is reduced as compared with the conventional example, and the moving image becomes smoother, so that the visibility can be improved.

(Second Embodiment) Hereinafter, a second embodiment of the present invention will be described with reference to FIGS. FIG. 3 shows a main block circuit of the MPEG video decoder 11 of the present embodiment. In this embodiment, the same components as those in the first embodiment have the same reference numerals, and a detailed description thereof will be omitted.

The MPEG video decoder 11 comprises a bit buffer 2, a frame buffer 22, a picture header detection circuit 3, a decode core circuit 4, and an underflow control circuit 12. The decoding result (video output) of each picture generated by the decoding core circuit 4 is transferred to each area 22a to 22c of the frame buffer 22. Further, each area 22a to 22 of the frame buffer 22
The decoding result of each picture read from c is transferred to the decoding core circuit 4.

The frame buffer 22 comprises a RAM,
The inside is divided into three areas (a forward reference area 22a, a backward reference area 22b, and a B picture storage area 22c). In the forward reference area 22a, a decoding result (video output) of a future I picture or P picture used when performing backward prediction in the decoding core circuit 4 is stored. The decoding result of the past I picture or P picture used when performing forward prediction in the decode core circuit 4 is stored in the backward reference area 22b.
The decoding result of the B picture is stored in the B picture storage area 22c. Then, the video output stored in any one of the areas 22a to 22c is output to a display (not shown).

Frame buffer 22 and bit buffer 2
Is an MPEG video decoder 1 with a reduced number of parts.
In order to reduce the cost of one, the area is provided separately in one RAM. By the way, the forward reference area 22a
Since the I picture or the P picture stored in the backward reference area 22b is used as the base data for performing the forward prediction or the backward prediction, the I picture or the P picture must be continuously stored in each area 22a, 22b until it becomes unnecessary. No. The B picture is not treated as the base data, and is unnecessary when output to the display 8.
Each of the regions 22a to 22c is called a plane.

When the MPEG video decoder and the MPEG audio decoder are mounted on one LSI, the MPEG audio decoder bit buffer (audio bit buffer) is also used as the MPEG video decoder frame buffer 22 and the bit buffer. (Video bit buffer) 2 and an area are provided separately in one RAM. For example, when a video CD is used as the transmission medium, a 4 MDRAM is used, the capacity of the video bit buffer 2 is 54 kbytes, the capacity of each of the areas 22 a to 22 c of the frame buffer 22 is 148.5 kbytes, and the capacity of the audio bit buffer is The capacity is set to 6.5 kbytes, and the capacity of the user area is set to 8 kbytes. By the way, the user area is a video CD
It is used for v2.0 standard sector buffers.

The underflow control circuit 12 compares the occupation amount Bm of the bit buffer 2 with the third threshold value BTH3, and detects whether the bit buffer 2 underflows. Here, the third threshold BTH3 is set to VD (Vbv [Video Buffering Verifier] De
lay) (BTH3 = RB x V)
D). Note that VD is defined by a picture header. The underflow control circuit 12 performs a decoding operation of the decode core circuit 4 and a read operation of a picture from the bit buffer 2 based on the type of the picture read from the bit buffer 2 and the comparison result and the detection result. And control.

Next, the operation of this embodiment will be described with reference to the flowchart shown in FIG. First, in S11,
The underflow control circuit 12 determines that the occupation amount Bm is the third
Is determined to be below the threshold value BTH3 (Bm <
(BTH3), the process proceeds to S12, and if it is determined that the value is not below (Bm ≧ BTH3), the process proceeds to S13. Here, when the occupation amount Bm is smaller than the third threshold value BTH3, when the next picture is read from the bit buffer 2, there is a high possibility that an underflow will occur.

At S12, error processing is performed.
That is, the underflow control circuit 12 stops reading the picture from the bit buffer 2. At the same time, the underflow control circuit 12 continues to output the video output, which is the decoding result of the picture previously read from the bit buffer 2, from the decode core circuit 4 instead of the picture currently being processed (repeat). ). Then, the process returns to S11.

In S13, the next picture is read from the bit buffer 2. Then, the decode core circuit 4 decodes the picture to generate a video output. Then, control goes to a step S14. In S14, when the underflow control circuit 12 determines that the bit buffer 2 does not underflow, the process returns to S11, and when it is determined that the underflow has occurred, the process proceeds to S15. That is, the decode core circuit 4
If the decoding process for one picture is completed normally, the process returns to S11. If the bit buffer 2 underflows during the decoding process for one picture, the process returns to S15.

In S15, the same error processing as in S12 is performed. Then, control goes to a step S16. In S16, the picture header detection circuit 3 determines the type of the picture read from the bit buffer 2, and if the picture is a B picture, the procedure proceeds to S17, and if the picture is an I picture or a P picture, the procedure proceeds to S18. .

In S17, the B picture that has been partially decoded by the decode core circuit 4 is skipped. Then, the process returns to S11. In S18,
The bit buffer 2 is controlled by the underflow control circuit 12.
If it is determined that the underflow has been released, S
Return to 13. That is, the process returns to S13 after waiting until a new video stream is transferred from the transmission medium and the underflow of the bit buffer 2 is released.

As described above, according to the present embodiment, the following operations and effects can be obtained. When the occupation amount Bm of the bit buffer 2 falls below the third threshold value BTH3 (that is, when there is a high possibility that underflow occurs), error processing is performed. as a result,
The occupation amount Bm of the bit buffer 2 increases, and the underflow hardly occurs.

Error processing is also performed when the bit buffer 2 underflows. As a result, even when the bit buffer 2 underflows, the video output from the decode core circuit 4 to the display is continuously output to the display without interruption, and the screen display on the display is also continued. When bit buffer 2 underflows,
If the decoding core circuit 4 is in the process of decoding a B picture, the B picture is skipped. Here, as described above, a B picture is generated by bidirectional prediction, and its importance is lower than that of an I picture or a P picture. Therefore, even if the B picture read from the bit buffer 2 is skipped, the decoding process can be performed in the decode core circuit 4 on the next picture read from the bit buffer 2 irrespective of its type.

When the bit buffer 2 underflows, if the I / P picture is being decoded in the decode core circuit 4,
After waiting until the underflow of the bit buffer 2 is released, the remaining decoding processing is continued. Therefore, I pictures and P pictures having high importance can be effectively used.

As described above, the number of dropped frames occurring in a moving image reproduced on the display is reduced as compared with the conventional example, and the moving image becomes smoother, so that the visibility can be improved. In the present embodiment, three areas 2 are stored in the frame buffer 22.
The effect described above can be obtained by applying the present invention when only 2a to 22c are provided. Therefore, when the frame buffer 22 has an area dedicated to output to a display in addition to the three areas 22a to 22c, it is not necessary to apply this embodiment.

As described above, when the frame buffer 22 and the bit buffer 2 are provided in one 4MDRAM, the frame buffer 22 has three regions 22a to 22a.
Only a capacity of 22c can be secured. Therefore, a B picture (hereinafter referred to as a second B
Picture) and decode the result to B
During the transfer to the picture storage area 22c, a B picture (hereinafter, referred to as a first B picture) already stored in the B picture storage area 22c is output to the display. As a result, the second B
When the picture is being decoded, the first B picture already stored in the B picture storage area 22c is newly decoded by the decode core circuit 4 into the second B picture.
B picture is overwritten.

In other words, when the bit buffer 2 underflows and the decoding operation in the decode core circuit 4 is interrupted, the second B picture in the middle of decoding and the first non-overwritten first B picture are stored in the B picture storage area 22c. The B picture and the B picture coexist. as a result,
The display screen of the display is divided into a previous screen and a screen being decoded. However, if the B picture being decoded is skipped as described above, such division of the screen is avoided.

By the way, when an I picture or a P picture is decoded by the decode core circuit 4 and the decoding result is transferred to the forward reference area 22a, it is already stored in the display in the backward reference area 22b or the B picture storage area. Is output. When an I picture or a P picture is decoded by the decode core circuit 4 and the decoded result is transferred to the backward reference area 22b, the display is displayed on the forward reference area 22b.
The picture already stored in the a or B picture storage area is output. Therefore, I
When decoding a picture or a P picture, the problem as in the case of a B picture does not occur.

Therefore, even if the bit buffer 2 underflows, if a new video stream is transferred from the transmission medium and the underflow of the bit buffer 2 is released, the I-picture or the P-picture becomes effective. You can make use of it. That is, for the I picture or P picture that is being decoded at the time when the bit buffer 2 underflows, the remaining decoding processing is continuously performed after the underflow is released.
As a result, the forward reference area 22a or the backward reference area 22
b can store an I picture or a P picture that has been completely decoded.

The above embodiments may be modified as follows, and the same operation and effect can be obtained in such a case. (1) The first embodiment and the second embodiment are used together. In this case, the effects of both embodiments can be provided. (2) In the first embodiment, the second threshold value BTH2 and the operation related thereto are omitted. In this case, the first threshold BTH
Actions and actions relating to 1 can be obtained.

(3) The first and second embodiments are replaced with software processing using a CPU. That is,
The signal processing in each of the circuits (3 to 6, 12) is replaced with software signal processing using a CPU. Although the embodiments have been described above, technical ideas other than the claims that can be grasped from the embodiments will be described below along with their effects.

(A) The MPEG video decoder according to any one of claims 1 to 5 , wherein the bit buffer comprises a RAM having a FIFO structure. In this way, writing and reading of the MPEG video stream can be performed easily. (B) In the MPEG video decoder according to any one of claims 1 to 5 , the circuit excluding the bit buffer has one circuit.
MPEG video decoder formed on a chip.

In this way, the size of the MPEG video decoder can be reduced. By the way, in this specification, the members according to the configuration of the present invention are defined as follows. (A) The bit buffer control means comprises an overflow determination circuit 5 and a picture skip circuit 6.

(B) The outside refers to a transmission medium, which includes a communication medium such as a LAN (Local Area Network), a video CD and a DVD (Digital Video Dis
k) and storage media such as VTR (Video Tape Recorder), terrestrial broadcasting and satellite broadcasting, and CATV (Community
broadcasting media, such as ity Antenna Television).

[0072]

According to the present invention, it is possible to provide an MPEG video decoder which can reduce the number of dropped frames in a reproduced moving image and improve the visibility.

[Brief description of the drawings]

FIG. 1 is a block circuit diagram of a first embodiment.

FIG. 2 is a flowchart for explaining the operation of the first embodiment.

FIG. 3 is a block circuit diagram of a second embodiment.

FIG. 4 is a flowchart for explaining the operation of the second embodiment.

FIG. 5 is a block circuit diagram of a conventional example.

FIG. 6 is a graph for explaining a conventional example.

[Explanation of symbols]

 2 bit buffer 3 picture header detection circuit 4 MPEG video decode core circuit 5 overflow determination circuit 6 picture skip circuit 12 underflow control circuit

──────────────────────────────────────────────────続 き Continuation of front page (56) References JP-A-7-135659 (JP, A) JP-A-6-292187 (JP, A) JP-A-6-261303 (JP, A) JP-A-6-261303 153184 (JP, A) (58) Field surveyed (Int. Cl. ⁷ , DB name) H04N ^7/ 24-7/68

Claims (5)

(57) [Claims] 1. A bit buffer for storing a video stream transferred from the outside and reading the video stream by one picture, and a picture buffer of the picture based on a picture header of the video stream read from the bit buffer. A picture header detection circuit for judging the type; an MPEG video decoding core circuit for decoding each picture in accordance with the MPEG video part; an overflow judgment for detecting the occupancy of the bit buffer and comparing the occupancy with a threshold value Circuit, when the occupancy of the bit buffer exceeds a threshold, when an I picture or a P picture is read from the bit buffer, the picture is directly transferred to the MPEG video decoding core circuit, and when a B picture is read, the corresponding picture is read. Picture the MP MPEG video decoder, characterized in that a picture skip circuit to be skipped without transferring the G video decode core circuit. 2. A bit buffer in which video streams transferred from the outside are sequentially stored, and the video stream is read out for one picture at regular intervals, and a picture header is stored from the video streams stored in the bit buffer. A picture header detection circuit for detecting and determining the type of a picture read from the bit buffer based on the picture header; and decoding each picture in accordance with the MPEG video part to output a video output for each picture. MPE to generate
A G video decode core circuit; an overflow determination circuit for detecting the occupancy of the bit buffer and comparing the occupancy with a threshold; and a first and a second node, which are connected to the first node. And each picture read from the bit buffer is transferred to the MPEG video decoding core circuit as it is, and when connected to a second node, each picture read from the bit buffer is transferred to the MPEG video decoding core circuit. A picture skipping circuit for skipping in units of pictures without transferring the picture to the picture buffer. The picture skipping circuit, when the occupation amount of the bit buffer exceeds a threshold, when an I picture or a P picture is read from the bit buffer. Connected to the first node side, and when the B picture is read out, MPEG video decoder characterized in that it is continued. 3. The MPEG video decoder according to claim 1 or 2, the capacity of the bit buffer (B
A) is set to a value obtained by adding a value (X) obtained by dividing a bit rate by a picture rate to a capacity (B) defined according to a buffer size (Vbv Buffer Size). An MPEG video decoder characterized in that a buffer size, a bit rate, and a picture rate are respectively specified by a sequence header attached to a head of a video stream sequence. 4. The MPEG video decoder according to claim 1 or 2, the capacity of the bit buffer (B
A) is set to exceed a value obtained by adding a value (X) obtained by dividing a bit rate by a picture rate to a capacity (B) defined according to a buffer size (Vbv Buffer Size). An MPEG video decoder, wherein the buffer size, bit rate, and picture rate are respectively specified by a sequence header attached to the head of a video stream sequence. In MPEG video decoder according to 5. A method according to claim 3 or 4, wherein the threshold MPEG video decoder, characterized in that it is set to the same value as the capacitance (B).