JP2009194591A - Reencoding apparatus, video recording apparatus, integrated circuit, and reencoding method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve responsiveness of a reencoding apparatus while preventing the deterioration in image quality. <P>SOLUTION: A decoding processor 101 obtains a decoding video data stream by decoding an encoding video data stream, and obtains an image type of each frame. A GOP structure determination part 105 determines an image type of each frame indicated by the decoding video data stream. An image coding processor 106 performs a first encoding process using the image type obtained by the decoding processor 101 and a second encoding process using the image type determined by the GOP structure determination part 105 on the decoding video data stream. An encoding controller 104 causes the image coding processor 106 to perform the first encoding process when the decoding video data stream includes the head of the GOP including the frame, and to perform the second encoding process when the decoding video data stream does not includes the head of the GOP. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a re-encoding device, a video recording device, an integrated circuit, and a re-encoding method, and in particular, decodes an encoded video data stream as digital video data into a decoded video data stream, and the decoded video data The present invention relates to a video encoding technique for encoding a stream.

  Conventionally, a video recording device such as a DVD recorder is used to decode a coded video data stream to obtain a decoded video data stream for the purpose of changing a data compression rate or format, and re-decode the decoded video data stream. A re-encoding device for encoding may be provided.

  In such a re-encoding device, image quality degradation may occur because the same image is regarded as a different image type at the time of initial encoding and at the time of re-encoding. For example, in a re-encoding device that performs encoding in accordance with the MPEG (Moving Picture Experts Group) system, a decoded video data stream obtained by decoding an input encoded video data stream is a GOP (Group Of Picture). It is data for one, I (Intra) picture, P (Predictive) picture, B (Bidirectionally predictive) picture are arranged in order of B0, B1, I2, B3, B4, P5, B6, B7, P8 Suppose that In this case, the decoded video data stream is regarded as data having a picture type (image type) different from that at the time of the first encoding, for example, B0, I1, B2, B3, P4, B5, B6, and P7, and is re-encoded. When the image is converted, a B picture having a high degree of image quality deterioration is used as a reference image at the time of re-encoding, and as a result, the re-encoding accuracy is lowered and a large image quality deterioration occurs.

  Patent Document 1 discloses a technique for matching image types before and after re-encoding in order to reduce the image quality degradation as described above.

Japanese Patent Application Laid-Open No. 2004-228620 obtains an encoded video data stream by encoding a decoded video data stream, detects the head of a GOP (image group) from the encoded video data stream, and detects the encoded video data stream. Is disclosed from the head of the GOP.
JP 2000-333163 A JP 2003-199021 A

  However, in the re-encoding device, when the re-encoded video data stream after re-encoding is output from the top of the image group as in Patent Document 2, when the number of images included in one image group is large, The time from when the encoded video data stream is input to the re-encoding device to when the re-encoded video data stream is output becomes longer, resulting in poor responsiveness.

  On the other hand, in order to improve the responsiveness of the re-encoding device, it can be considered that the decoded video data stream is regarded as a different image type from the initial encoding and re-encoding is performed from the middle of the image group. In such a case, an image type image having a large degree of image quality degradation may be used as a reference image, which may cause a large image quality degradation.

  An object of the present invention is to improve the responsiveness of a re-encoding device while suppressing deterioration in image quality in view of the above points.

In order to solve the above problem, the re-encoding process of the present invention
A decoding process for decoding the encoded video data stream to obtain a decoded video data stream, and obtaining an image type of each image indicated by the decoded video data stream;
An image group configuration process for determining an image type of each image indicated by the decoded video data stream obtained by the decoding process;
For each image indicated by the decoded video data stream obtained by the decoding process, it is determined whether or not the decoded video data stream includes a head of a predetermined specification image group including the image, and the decoding When the video data stream includes the head, the decoded video data stream is subjected to a first encoding process for encoding the image by regarding the image as the image type acquired by the decoding process, A second encoding that encodes the decoded video data stream by regarding the decoded video data stream as an image type determined by the image group configuration process when the decoded video data stream does not include the head; And an image encoding process for performing the process.

  Accordingly, when the decoded video data stream is acquired from the middle of the image group by the decoding process, the second encoding process, that is, the image group configuration process is performed until the decoded video data stream at the head of the image group is acquired. After the encoding using the image type determined by the above is performed and the decoded video data stream at the head of the image group is acquired, the first encoding process, that is, the image type acquired by the decoding process is used. Encoding is performed. Therefore, since the re-encoded video data stream can be output from the middle of the image group, the output of the re-encoded video data stream can be started earlier than when the re-encoded video data stream is output from the head of the image group. Can be improved. In addition, after the decoded video data stream at the head of the image group is acquired, encoding is performed using the image type acquired by the decoding process, so that image quality deterioration can be suppressed.

  According to the present invention, responsiveness is improved and image quality deterioration is suppressed.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  As shown in FIG. 1, a re-encoding device 100 according to an embodiment of the present invention includes a decoding processing unit 101, a memory 102, an encoding parameter generation unit 103, an encoding control unit 104, and a GOP (Group Of Pictures) structure determination. Unit (image group configuration unit) 105 and an image encoding processing unit 106.

  The decoding processing unit 101 acquires (generates) a decoded video data stream by decoding the encoded video data stream encoded by the MPEG2 system, and stores the acquired decoded video data stream in the memory 102. The decoding processing unit 101 performs a picture type (picture_coding_type) indicating whether a frame (image) indicated by the acquired decoded video data stream is an I picture, a P picture, or a B picture during the decoding process. Get (acquire) parameters such as vector (motion_vector).

  The encoding parameter generation unit 103 acquires the parameter obtained by the decoding processing unit 101, and generates an encoding parameter based on the parameter. Specifically, the acquired parameters are rearranged in the encoding process order and output as encoding parameters.

  When the decoding video data stream is acquired by the decoding processing unit 101, the encoding control unit 104 uses the picture type obtained by the decoding processing unit 101 to set the GOP reconfiguration flag corresponding to each frame to ON or Set to OFF. Specifically, as shown in FIG. 2, the encoding control unit 104 first acquires the picture type of the frame for which the GOP reconstruction flag is set in S201, and in S202, the picture type acquired in S201 is an I picture. It is determined whether or not. If it is determined in S202 that the picture type is I picture, in S203, the GOP reconstruction flag for two consecutive frames that are B pictures before the frame (I picture) for which the GOP reconstruction flag is set is set to OFF. In step S205, the GOP reconstruction flag of the GOP reconstruction flag setting target frame (I picture) is set to OFF. If it is determined in S202 that the picture type is not an I picture, it is determined in S204 whether or not the GOP reconstruction period has ended. Specifically, if the GOP reconstruction flag of the previous frame is set to OFF, it is determined that the GOP reconstruction period has ended, and the previous frame does not exist or the GOP reconstruction flag of the previous frame is set. If it is set to ON, it is determined that the GOP reconstruction period has not ended. If it is determined in S204 that the GOP reconfiguration period has ended, the GOP reconfiguration flag is set to OFF in S205, and if it is determined in S204 that the GOP reconfiguration period has not ended, in S206 Set the GOP reconfiguration flag to ON.

  Then, when the image encoding processing unit 106 re-encodes the decoded video data stream of the frame whose corresponding GOP reconstruction flag is ON, the encoding control unit 104 causes the GOP structure determining unit 105 to Start.

  When activated by the encoding control unit 104, the GOP structure determination unit 105 determines the picture type of each frame indicated by the decoded video data stream acquired by the decoding processing unit 101.

  Specifically, the GOP structure determination unit 105 determines the picture type for each frame according to the procedure shown in FIG. In step S301, the GOP structure determination unit 105 first determines whether three frames ahead of the picture type determination target frame are in the GOP reconstruction period. That is, it is determined whether the GOP reconstruction flag of the third frame ahead is ON or OFF. If it is determined in S301 that the third frame ahead of the picture type determination target is not a GOP reconstruction period, the picture type of the frame is determined to be a P picture in S302. On the other hand, when it is determined in S301 that the third frame ahead of the picture type determination target is the GOP reconstruction period, it is determined in S303 whether the picture type of the frame is an I picture. If it is determined in S303 that the picture type is an I picture, the picture type of the frame is determined to be an I picture in S304. On the other hand, if it is determined in S303 that the picture type is not an I picture, it is determined in S305 whether or not two frames have already been determined as B pictures since the previous frame was determined as an I picture or P picture. If it is determined in S305 that two frames have already been determined as B pictures, the picture type of the frame is determined as P picture in S306. On the other hand, if it is determined in S305 that the two frames are not yet determined as B pictures, the picture type of the frames is determined as B pictures in S307.

  Note that the picture type determination operation in the procedure as shown in FIG. 3 is performed at the timing of performing the re-encoding process. This is because the GOP end picture type cannot be correctly determined if the decoding processing unit 101 obtains the picture type. That is, the determination in S301 cannot be performed.

  In addition, as described above, the GOP structure determination unit 105 has two or more frames between the first frame of the next GOP and less than two frames with the previous I picture or P picture. The picture type of a frame that is not an I picture is determined to be a B picture, but the number of frames between the first frame of the next GOP is a predetermined number other than two, The number of frames from (in between) the I picture or P picture may be less than the predetermined number, and the picture type of a frame that is not an I picture may be determined as a B picture. For example, the picture type of a frame that is 3 or more frames from the first frame of the next GOP, is less than 3 frames from the previous I picture or P picture, and is not an I picture May be determined as a B picture. In this case, it is determined in step S301 whether or not four frames ahead of the picture type determination target frame are in the GOP reconstruction period, and in step S305, the B picture has already been obtained for three frames since the frame was determined to be the previous I picture or P picture. It may be determined whether or not it has been determined.

  Further, when P pictures continue at the GOP end, a B picture may be sandwiched between them.

  The image encoding processing unit 106 reads out the decoded video data stream acquired by the decoding processing unit 101 from the memory 102, and regards each frame as the picture type acquired by the decoding processing unit 101 for the decoded video data stream. The first encoding process to be encoded and the second encoding process for encoding the decoded video data stream by regarding each frame as the picture type determined by the GOP structure determination unit 105 Obtain an encoded video data stream. In addition, when performing the first encoding process, the image encoding processing unit 106 generates a motion vector by performing a motion search using the motion vector acquired by the decoding processing unit 101, while the second encoding process is performed. When performing the encoding process, the image encoding processing unit 106 performs a motion search using a motion vector generated in the past to generate a motion vector.

  Then, when the image encoding processing unit 106 re-encodes the decoded video data stream of the frame whose corresponding GOP reconstruction flag is OFF, the encoding control unit 104 When the image encoding processing unit 106 performs re-encoding on a decoded video data stream of a frame whose corresponding GOP reconfiguration flag is ON while causing the first encoding process to be performed on the The encoding processing unit 106 performs the second encoding process.

  That is, for each frame indicated by the decoded video data stream acquired by the decoding processing unit 101, the encoding control unit 104 determines whether or not the decoded video data stream includes the head of the GOP including the frame. If the decoded video data stream includes the head, the image encoding processing unit 106 performs a first encoding process, while the decoded video data stream includes the head. If not, the image encoding processing unit 106 is caused to perform the second encoding process.

  Here, referring to FIG. 4, I picture such as I2, B0, B1, P5, B3, B4, P8, B6, B7, P11, B9, P10, P14, B12, B13, I17, B15, B16 A case where an encoded video data stream in which a P picture and a B picture are arranged is input to the re-encoding device 100 will be described. In this case, the decoding processing unit 101 decodes the encoded video data stream, and B0, B1, I2, B3, B4, P5, B6, B7, P8, B9, B10, P11, B12, B13, P14, A decoded video data stream having a sequence of B15, B16, and I17 is acquired and stored in the memory 102. Since this decoded video data stream starts (exists) from the beginning of the GOP, the image encoding processing unit 106 performs the first encoding process from the first frame in accordance with the control by the encoding control unit 104. . That is, the decoded video data stream is encoded by regarding the picture type of each frame as not changing from the picture type obtained by the decoding processing unit 101. And as a result, re-encoded video data stream of I2, B0, B1, P5, B3, B4, P8, B6, B7, P11, B9, P10, P14, B12, B13, I17, B15, B16 is obtained. To do.

  Next, referring to FIG. 5, B4, P8, B6, B7, P11, B9, B10, P14, B12, B13, I17, B15, B16, etc. A case where the encoded video data stream is input to the re-encoding device 100 will be described. In this case, the decoding processing unit 101 decodes the encoded video data stream and arranges B4, P5, B6, B7, P8, B9, B10, P11, B12, B13, P14, B15, B16, and I17. A decoded video data stream is obtained and stored in the memory 102. Since the GOP head frame (picture) is missing in this decoded video data stream, the image encoding processing unit 106 performs the second encoding process in accordance with the control by the encoding control unit 104. That is, each frame is regarded as a picture type newly determined by the GOP structure determination unit 105, and is called I2, B0, B1, P5, B3, B4, P8, B6, B7, P9, P10, I13, B11, B12 The decoded video data stream is encoded so as to obtain a sequence of re-encoded video data streams. However, from B15, B16, and I17 constituting the head of the next GOP, each frame indicated by the decoded video data stream is regarded as the same picture type as that at the time of decoding and is re-encoded (first encoding process). I do.

  Even when the decoded video data stream starts from the beginning of the GOP, there is a syntax error in the encoded video data stream, and the decoding processing unit 101 cannot correctly acquire parameters such as the picture type. If the encoded video data stream cannot be decoded correctly, the encoding control unit 104 may cause the image encoding processing unit 106 to perform the second encoding process. When there is a syntax error, for example, a sequence header including an image size (horizontal_size_value, vertical_size_value), an extension header including progressive information (progressive_sequence, picture_structure, progressive_frame), or a picture including a picture type (picture_coding_type) This is the case when an error is detected in the header or the motion vector is incorrect. When there is a syntax error, there is a high possibility that the picture type obtained by the decoding processing unit 101 is incorrect, so that the encoding process is performed using the picture type newly determined by the GOP structure determining unit 105. By doing so, image quality deterioration can be prevented.

  Further, in this embodiment, the encoding control unit 104 detects, for each frame, whether or not the decoded video data stream acquired by the decoding processing unit 101 includes the head of the GOP including the frame. However, the determination may be made by detecting a GOP header (group_of_picture_header).

  In addition, the present invention is not limited to a re-encoding device that performs decoding processing and encoding processing that conforms to the MPEG2 format, but also performs re-processing that performs decoding processing and encoding processing that conforms to other encoding schemes such as the H.264 format. The present invention can also be applied to an encoding device. For example, when the present invention is applied to a re-encoding device that supports the H.264 scheme, a slice type (slice_type) can be used as an image type instead of a picture type, and an image group can be used instead of a GOP. A plurality of images of a predetermined specification starting from an I slice in the encoded state can be used. In the H.264 re-encoding apparatus, the encoding control unit 104 includes, for each frame, the decoded video data stream acquired by the decoding processing unit 101 includes the head of the image group including the frame. It can be determined by detecting an SPS (seq_parameter_set_rbsp) or IDR picture.

  FIG. 6 is a block diagram showing a configuration of a DVD recorder system 600 provided with the re-encoding device 100.

  The DVD recorder system 600 includes a receiving unit 601, a system LSI (Large Scale Integration) 602, a DRAM 603, and a storage interface unit (Storage Interface Unit) 604. The television 605 has a decoded video data stream and a decoded audio data stream. Is output.

  The receiving unit 601 receives radio waves, terrestrial waves, and signals from cables received by the satellite broadcasting antenna, and based on these radio waves and signals, an encoded video data stream and encoded audio data that are digital signals. Generate and output a stream. Specifically, the reception unit 601 includes an IQ detection unit (In-Phase & Quadrature detection unit) 601a, a QPSK (Quadrature Phase Shift Keying) demodulation LSI 601b, a down converter 601c, an AD converter 601d, a down converter 601e, a QAM (Quadrature Amplitude Modulation). A demodulation LSI 601f is provided. The radio wave received by the satellite broadcasting antenna is converted into an encoded video data stream and an encoded audio data stream by the IQ detector 601a and the QPSK demodulation LSI 601b, and the terrestrial wave is encoded video by the down converter 601c and the AD converter 601d. The data stream and the encoded audio data stream are converted, and the signal from the cable is converted into the encoded video data stream and the encoded audio data stream by the down converter 601e and the QAM demodulation LSI 601f.

  The system LSI 602 includes a decoder 602a, an encoder 602b, and a microcomputer 602c.

  The decoder 602a has the decoding processing unit 101 of the re-encoding device 100, and acquires the decoded video data stream by decoding the encoded video data stream output by the receiving unit 601 by the function of the decoding processing unit 101. (Generate). Also, the decoder 602a decodes the encoded audio data stream to obtain a decoded audio data stream. The decoder 602 a stores the acquired decoded video data stream and decoded audio data stream in the DRAM 603. The DRAM 603 includes the memory 102 of the re-encoding device 100.

  The encoder 602b includes an encoding parameter generation unit 103, an encoding control unit 104, a GOP structure determination unit 105, and an image encoding processing unit 106 of the re-encoding device 100, and performs decoding from the DRAM 603 by the functions of these units. A re-encoded video data stream is obtained by reading and encoding the video data stream. In addition, the encoder 602b obtains a re-encoded audio data stream by reading and decoding the decoded audio data stream from the DRAM 603.

  The microcomputer 602c controls the decoder 602a and the encoder 602b.

  The storage interface unit 604 records the re-encoded video data stream acquired by the image encoding processing unit 106 of the encoder 602b and the re-encoded audio data stream acquired by the encoder 602b on the DVD 606.

  According to the above DVD recorder system 600, when the decoded video data stream is acquired from the middle of the GOP by the decoding processing unit 101, it is determined by the GOP structure determining unit 105 until the decoded video data stream at the head of the GOP is acquired. Encoding is performed using the selected picture type, and after the decoded video data stream at the head of the GOP is acquired, encoding is performed using the picture type acquired by the decoding processing unit 101. Therefore, since the re-encoded video data stream can be recorded from the middle of the GOP, the recording of the re-encoded video data stream on the DVD 606 can be started earlier than when starting from the GOP head, and the responsiveness is improved. it can. In addition, since the decoding is performed using the picture type acquired by the decoding processing unit 101 after the GOP head decoded video data stream is acquired, it is indicated by the re-encoded video data stream recorded on the DVD 606. Degradation of the image quality is suppressed.

  Instead of the storage interface unit 604, a storage interface unit that records the re-encoded video data stream on a recording medium other than DVD, such as a hard disk or an SD memory card, may be provided.

  Further, the processing unit other than the memory 102 of the re-encoding device 100 may not necessarily be configured by the one-chip system LSI 602, and a part of the functions may be provided in another chip. Further, like the system LSI 602, a block having a function other than the re-encoding for the video data stream (for example, the re-encoding function for the audio data stream) may be provided in the system LSI including the re-encoding device 100. Good.

  The re-encoding device 100 is not limited to the DVD recorder system, and may be provided in other video recording devices such as a storage / playback device such as a video camera having a dubbing function, a digital television having a recording function, and the like.

  The re-encoding device, the video recording device, the integrated circuit, and the re-encoding method according to the present invention have an effect that the responsiveness is improved and the deterioration of the image quality is suppressed. For example, the encoding as digital video data This is useful as a video encoding technique for decoding a decoded video data stream into a decoded video data stream and encoding the decoded video data stream.

It is a block diagram which shows the structure of the re-encoding apparatus 100 which concerns on embodiment of this invention. 3 is a flowchart showing the operation of the encoding control unit 104. 3 is a flowchart showing the operation of the GOP structure determination unit 105. FIG. 4 is an explanatory diagram illustrating an example of an encoded video data stream, a decoded video data stream, and a re-encoded video data stream. FIG. 4 is an explanatory diagram illustrating an example of an encoded video data stream, a decoded video data stream, and a re-encoded video data stream. 2 is a block diagram showing a configuration of a DVD recorder system 600 including the re-encoding device 100. FIG.

Explanation of symbols

100 Re-encoding device 101 Decoding processing unit 104 Encoding control unit 105 GOP structure determination unit (image group configuration unit)
106 Image encoding processing unit 600 DVD recorder system (video recording apparatus)
601 Receiving unit 604 Storage interface unit

Claims (7)

A decoding processor that decodes the encoded video data stream to obtain a decoded video data stream, and obtains an image type of each image indicated by the decoded video data stream;
An image group configuration unit that determines an image type of each image indicated by the decoded video data stream acquired by the decoding processing unit;
A first encoding process for encoding each image as an image type acquired by the decoding processing unit for the decoded video data stream acquired by the decoding processing unit, and each image as the image group configuration unit An image encoding processing unit that performs a second encoding process for encoding the image type determined by
For each image indicated by the decoded video data stream acquired by the decoding processing unit, it is determined whether or not the decoded video data stream includes a head of an image group of a predetermined specification including the image, and the decoding When the encoded video data stream includes the head, the image encoding processor performs the first encoding process, while when the decoded video data stream does not include the head, A re-encoding apparatus comprising: an encoding control unit that causes the image encoding processing unit to perform a second encoding process. The re-encoding device according to claim 1, wherein
The decoding processing by the decoding processing unit, and the first and second encoding processing by the image encoding processing unit are based on the MPEG system,
The image type is a picture type;
The image group is a GOP;
In the image group configuration unit, the number of images between the first image of the next GOP is a predetermined number or more, the number of images from the previous I picture or P picture is less than the predetermined image, and is not an I picture A re-encoding device, wherein a picture type of an image is determined to be a B picture. The re-encoding device according to claim 1, wherein
The decoding processing unit acquires a motion vector in addition to the image type of each image at the time of decoding,
When performing the first encoding process, the image encoding processing unit generates a motion vector by performing a motion search using the motion vector acquired by the decoding processing unit. A re-encoding device, wherein when performing an encoding process, a motion vector is generated by performing a motion search using a motion vector generated in the past by the image encoding processing unit. The re-encoding device according to claim 1, wherein
When there is a syntax error in the encoded video data stream and the decoding processing unit cannot correctly decode the encoded video data stream, the encoding control unit causes the image encoding processing unit to A re-encoding apparatus that performs the encoding process of 2. A re-encoding device according to claim 1;
A receiving unit that receives a radio wave or signal, generates and outputs the encoded video data stream based on the radio wave or signal;
A video recording apparatus comprising: a storage interface unit that records a re-encoded video data stream acquired by encoding processing by the image encoding processing unit on a recording medium.   An integrated circuit comprising the re-encoding device according to claim 1. A decoding process step of decoding the encoded video data stream to obtain a decoded video data stream and obtaining an image type of each image indicated by the decoded video data stream;
An image group configuration step for determining an image type of each image indicated by the decoded video data stream obtained by the decoding processing step;
For each image indicated by the decoded video data stream acquired by the decoding processing step, it is determined whether or not the decoded video data stream includes a head of a predetermined image group including the image, and the decoding If the encoded video data stream includes the head, a first encoding process is performed on the decoded video data stream by regarding the image as an image type acquired by the decoding processing step. On the other hand, if the decoded video data stream does not include the head, the decoded video data stream is encoded by regarding the image as the image type determined by the image group configuration step. And an image encoding processing step for performing the encoding processing. Re-encoding method.

Images