KR101625441B1 - Method and device for encoding multi-track video to scalable video using fast motion estimation - Google Patents

Abstract

The present invention can predict intra-layer and inter-layer motion at high speed by using motion information of existing multitrack video when converting two or more multitrack video compressed to original video into SVC video and encoding the same, And more particularly, to a method and apparatus for encoding multitrack video into scalable video using high-speed motion prediction, which is capable of speeding up a motion prediction time having the highest computational amount in video encoding to increase the conversion speed to scalable video .
The video encoding apparatus according to the present invention receives two adjacent video frames for multitrack video, extracts macroblock partition information and motion vector information through bitstream decoding, compares partition information of a macroblock to be encoded based on the extracted macroblock partition information and motion vector information, If motion information is the same, motion information is compared. If motion information is the same, encoding is performed in an inter-layer motion prediction mode. If the partition information is not the same or motion information is not the same, (Motion prediction mode) or an SVC intra prediction mode (intra prediction mode).

Description

[0001] The present invention relates to a method and apparatus for encoding multi-track video into scalable video using fast motion prediction,

The present invention relates to a video encoding technique for compressing a video size to provide multi-track video as scalable video, and more particularly, to a video encoding technique for compressing two or more multi-track videos compressed for original video into SVC Video Coding) By encoding motion information of an existing multitrack video, it is possible to predict intra-layer and inter-layer motion at high speed, The motion estimation time having the highest calculation amount in the video encoding can be increased and the conversion speed to the scalable video can be improved.

In order to service contents including video and audio received from a contents provider (CP) to a user terminal, basic resolution scaling, frame rate conversion, video / audio encoding (Vidoe / audio encoding) Metadata ingestion, packaging, and so on.

At this time, when the user performs an inustaining process in the content server, when a video having an error is served to the user terminal, a problem occurs when the user terminal plays the video. In order to prevent this, the final result is reproduced and finally checked by the person.

However, since the majority of the video provided to the user terminal by the content server is large or large, the process of confirming each user in order to perform ingesting is long and has limitations. Have been proposed recently.

In addition, the online video service allows users to consume video by streaming video over the Internet. At this time, adaptive video streaming technology that transmits and adjusts the amount of video data according to the user's network environment A technique for preventing a user from interrupting a video or watching a video with a broken image is becoming common. The current adaptive video streaming technology is mainly composed of a technique of creating a compressed video having various data sizes for one original video and selecting a compressed video suited to the user's network environment, You must ingest multiple compressed video (multi-track video).

In recent years, Scalable Video Coding (SVC) method has been developed by H.264 based on joint video technology (JVT) group of ITU and MPEG with the aim of providing video service for various devices and network environment with one compressed video .

However, since SVC has been standardized recently, it is in the initial stage of commercialization. Currently, SVC prepares video files by using existing coding method such as H.264, and provides files corresponding to each device and network environment And a multi-track video system. When switching to SVC in the future, it is necessary to re-encode the existing multi-track video into the SVC, and this re-encoding is time-consuming.

Korean Patent Publication No. 10-2011-0053245 (2011.05.19)

In order to solve the above-described problems, the present invention provides a method and apparatus for converting a multi-track video compressed to an original video into a Scalable Video Coding (SVC) video and encoding the multi- intra-layer and inter-layer motion can be predicted at a high speed using motion information, thereby speeding up the motion estimation time with the highest calculation amount in video encoding It is an object of the present invention to provide a method and apparatus for encoding multitrack video into scalable video using fast motion prediction, which can improve the conversion speed to scalable video.

A method of decoding an SVC video according to an exemplary embodiment of the present invention includes receiving a bitstream for video of a plurality of layers, encoding mode and prediction information for a macroblock of a lower layer among the plurality of layers, information on the macroblock of the upper layer, performing prediction on the macroblock of the lower layer using the determined encoding mode and prediction information, determining an encoding mode and prediction information for a macroblock of an upper layer among the plurality of layers And performing an inter-layer prediction on an upper layer video using the encoding mode and the prediction information of the upper layer macroblock to reconstruct the video.

In addition, the encoding mode information for the macroblock of the lower layer may indicate an inter prediction mode or an intra prediction mode.

In addition, the encoding mode information of the upper layer macroblock may indicate one of an inter prediction mode, an intra prediction mode, and an interlayer prediction mode.

In addition, the prediction information for the macroblock of the lower layer may include motion vector information and macroblock partition information.

According to the present invention, intra-layer and inter-layer motion prediction can be performed at high speed by utilizing motion information of existing multitrack video when converting multitrack video into scalable video.

In addition, the speed of conversion into scalable video can be improved by speeding up the motion prediction time having the highest calculation amount in video encoding.

In addition, it is possible to shorten the encoding time of the multi-track video into the scalable video through high-speed motion prediction in the scalable video, thereby enabling speeding up.

BRIEF DESCRIPTION OF DRAWINGS FIG. 1 is a block diagram schematically showing the overall configuration of a video providing system according to an embodiment of the present invention; FIG.
2 is a block diagram schematically illustrating an internal functional block of a video encoding apparatus according to an embodiment of the present invention.
3 is a flowchart illustrating a method of encoding multitrack video into scalable video using fast motion prediction of a video encoding apparatus according to an embodiment of the present invention.
4 is a diagram illustrating an example of generating an original video as N multitrack video according to an embodiment of the present invention.
5 is a flowchart illustrating a video encoding process using motion prediction according to an embodiment of the present invention.
6 is a diagram illustrating a process of encoding an Nth layer of an SVC encoding unit according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Reference will now be made in detail to the preferred embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the like elements throughout. Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

BRIEF DESCRIPTION OF DRAWINGS FIG. 1 is a block diagram schematically showing the overall configuration of a video providing system according to an embodiment of the present invention; FIG.

Referring to FIG. 1, a video providing system 100 according to the present invention includes a video encoding apparatus 110, a communication network 120, and a user terminal 130.

The video encoding apparatus 110 receives two adjacent video streams for multitrack video, extracts macroblock partition information and motion vector information through bitstream decoding, and generates a motion prediction mode or an SVC And performs encoding in an intra prediction mode (intra prediction mode).

Here, the video encoding apparatus 110 compares the partition information of the macroblock to be encoded, compares the motion information with the motion information, and encodes the motion information in an inter-layer motion prediction mode if the motion information is the same If the partition information is not identical or the motion information is not the same, encoding is performed in the encoding mode of the current macroblock.

The video encoding apparatus 110 transmits the re-encoded SVC video to the user terminal 130 via the communication network 120 and provides the re-encoded SVC video.

Here, the video encoding apparatus 110 may be a media server or the like that transmits the re-encoded SVC video to one or more user terminals 130. The original video may be input and may be ingested into two or more multitrack video A dedicated media device that can re-encode it as SVC video and transmit it over the communication network 120, and so on.

The communication network 120 provides a transmission path for transmitting video from the video encoding apparatus 110 to the user terminal 130 and provides a connection path for the user terminal 130 to access the video encoding apparatus 110. Here, the communication network 120 includes a mobile communication network such as WCDMA, HDPA, 3G, and 4G, a Bluetooth, a Zigbee, a Wi-Fi, and a wired communication network such as the Internet or PSTN .

The user terminal 130 receives the SVC video transmitted from the video encoding apparatus 110, decodes and displays the SVC video.

Here, the user terminal 130 may be an IPTV, a settop box, or the like capable of receiving video data from the video encoding apparatus 110 and displaying the video data. In addition, A smart phone or a mobile communication terminal.

2 is a block diagram schematically illustrating an internal functional block of a video encoding apparatus according to an embodiment of the present invention.

2, a video encoding apparatus 110 according to the present invention includes a communication unit 210, an original video storage unit 220, a multi-track video encoder 230, a video arrangement unit 240, An SVC encoding unit 260, and the like.

The communication unit 210 communicates with the user terminal 130 through the communication network 120.

That is, the communication unit 210 receives the transmission request of the contents including video and audio from the user terminal 130 through the communication network 120, or transmits the re-encoded SVC video to the user terminal 130.

The original video storage unit 220 stores a plurality of original videos including audio and video according to various types.

The multitrack video encoder 230 receives the original video and generates two or more N multitrack videos having different image qualities as shown in FIG. 4 is a diagram illustrating an example of generating an original video as N multitrack video according to an embodiment of the present invention.

The video arranging unit 240 arranges two or more multitrack videos according to SVC encoding in order of quality.

At this time, the video arranging unit 240 arranges video having a higher resolution for two or more multi-track videos on a higher layer and transmits a video having a higher frame rate to a higher layer And can align video with a higher bit rate to a higher layer if the resolution and frame rate are the same.

The bitstream analyzing unit 250 receives two adjacent video signals from the aligned multitrack video and extracts macroblock partition information and motion vector information through bitstream decoding To the SVC encoding unit 260.

In addition, the bitstream analyzing unit 250 obtains SVC (N-1) encoded by the SVC with the compressed video N having the highest image quality and the compressed video N-1 having the next higher image quality ) th layer bitstream, and transmits the macroblock partition information and the motion vector information to the SVC encoding unit 260. The SVC encoding unit 260 encodes the macroblock partition information and the motion vector information. The SVC encoding unit 260 performs a motion prediction mode or an SVC intra prediction mode on a macroblock to be encoded based on macroblock partition information and motion vector information received from the bitstream analyzing unit 250. [ (intra prediction mode).

Herein, the motion prediction mode refers to an encoding mode in which temporal redundancy of a macroblock is removed by using motion prediction, and the SVC intra prediction mode is an intra prediction mode of H.264 and an inter-layer intra-prediction Quot; mode " Intraprediction means a method of predicting the pixel value of the current block by using the pixle value around the block. Interlace intra prediction can predict the pixel value of the current block using the pixle value of the lower layer at this time . The inter-layer motion prediction mode refers to a case in which a motion vector of a lower layer is utilized in a motion vector prediction in an SVC.

In addition, the SVC encoding unit 260 can encode the " compressed video N "having the highest image quality among the two compressed videos into the Nth layer of the SVC.

In addition, the SVC encoding unit 260 determines whether the current macroblock to be encoded is encoded in a motion prediction mode, and encodes it into an SVC intra prediction mode if it is not a motion prediction mode In the case of the motion prediction mode, the motion information is compared between the lower layer and the upper layer. If motion information is the same, the motion information is encoded in an inter-layer motion prediction mode, If the two pieces of partition information are not the same or the two pieces of motion information are not the same, the encoding mode of the current macroblock can be encoded.

In this case, in the case of the motion prediction mode, the SVC encoding unit 260 outputs up-scaled information of a partition of a macroblock corresponding to an (N-1) th SVC layer bitstream corresponding to a lower layer, And compares the partition information of the macroblock to confirm that they are the same.

If the two pieces of partition information are the same, the SVC encoding unit 260 outputs up-scaled information on the motion information of the macroblock corresponding to the (N-1) th SVC layer bitstream corresponding to the lower layer The motion information of the current macroblock is compared to confirm that the motion information is the same.

The SVC encoding unit 260 includes up-scaled information of the motion information of the macroblock corresponding to the (N-1) th SVC layer bitstream corresponding to the lower layer up to a half pel unit, Only two motion information are compared to half pel to confirm whether two motion information are identical.

3 is a flowchart illustrating a method of encoding multitrack video into scalable video using fast motion prediction of a video encoding apparatus according to an embodiment of the present invention.

Referring to FIG. 3, a video encoding apparatus 110 according to the present invention inputs original video to a multitrack video encoder 230 to generate two or more multitrack videos (S310).

Then, the video encoding apparatus 110 arranges two or more multi-track videos in order of quality (Quality) through the video arranging unit 240 (S320).

Next, the video encoding apparatus 110 inputs two neighboring videos out of the aligned multitrack video to the bitstream analyzing unit 250 and performs bitstream decoding to extract macroblock partition information, And extracts vector information (S330).

At this time, the video encoding apparatus 110 generates SVC (N-1) which encodes the compressed video N having the highest image quality among the sorted multitrack video and the compressed video N-1 adjacent to the next highest quality, 1) th macroblock partition information and motion vector information for two videos of the th layer bit stream.

Here, the SVC encoding unit 260 can encode "compressed video N" of the two compressed videos into the Nth layer of the SVC.

Next, the video encoding apparatus 110 encodes the macroblock to be encoded into a motion prediction mode or an intra prediction mode (SVC) based on the macroblock partition information and motion vector information (S340).

That is, the video encoding apparatus 110 performs encoding in the motion prediction mode or the SVC intra prediction mode through the SVC encoding unit 260 as shown in FIG. 5 is a flowchart illustrating a video encoding process using motion prediction according to an embodiment of the present invention.

First, the video encoding apparatus 110 determines whether a current macroblock to be encoded is encoded in a motion prediction mode (S510). If it is not a motion prediction mode (No in S510), the video encoding apparatus 110 performs an intra prediction mode intra prediction mode) (S580).

Then, the video encoding apparatus 110 determines whether the macroblock corresponding to the (N-1) th SVC layer bitstream corresponding to the lower layer is the motion prediction mode in the case of the motion prediction mode (YES in step S510) (S520). If the mode is not the motion prediction mode (S520: NO), encoding is performed in the encoding mode of the current macroblock (S590).

However, when the macroblock corresponding to the (N-1) th SVC layer bitstream corresponding to the lower layer is the motion prediction mode (YES in step S520), the video encoding apparatus 110 performs up- The macroblock partition of the lower layer is compared with the macroblock partition of the upper layer (S530).

If the two partitions are not the same (S540-No), the video encoding apparatus 110 performs encoding in the encoding mode of the current macroblock (S590). If the two partitions are identical (S540-Yes) The motion vector of the layer is compared with the motion vector of the upper layer (S550).

If the two motion vectors are not equal to each other (S560: No), the encoding mode of the current macroblock is set to the encoding mode of the current macroblock (S590). If the two motion vectors are the same (S560-YES), the macroblock is encoded in the interlayer motion prediction mode (S570).

The video encoding apparatus 110 performs the entire SVC encoding by performing the above-described process while incrementing N by 1 to N = K when the number of multi-track videos is K at N = 2.

6 is a diagram illustrating a process of encoding an Nth layer of an SVC encoding unit according to an embodiment of the present invention.

Referring to FIG. 6, when an image is input, the SVC encoding unit 260 according to the present invention needs to determine an encoding mode for encoding an input image. Based on the lower layer texture information, Intra prediction mode or a motion prediction mode based on lower layer motion information and inverse quantization / IDCT. The SVC encoding unit 260 performs DCT (Discrete Cosine Transform) / Quantization process and Context-adaptive variable-length coding (CAVLC) / Context-adaptive binary arithmetic coding (CABAC) on the SVC video according to the determined operation mode. ) Process.

That is, the video encoding apparatus 110 encodes the video data into an SVC intra prediction mode based on the lower layer texture information through the SVC encoding unit 260, and generates a motion prediction mode (motion prediction mode), compares the partition information of the macroblock to be encoded, compares the motion information with the motion information, and encodes the motion information in an inter-layer motion prediction mode if motion information is the same And performs encoding in the encoding mode of the current macroblock if the partition information is not the same or the motion information is not the same.

As described above, according to the present invention, when converting two or more multi-track videos compressed to original video into SVC (Scalable Video Coding) video and encoding the same, motion of the existing multi- Information can be used to predict intra-layer and inter-layer motion at high speed, thereby speeding up the motion estimation time with the highest calculation amount in video encoding, It is possible to realize a method and apparatus for encoding multi-track video into scalable video using fast motion prediction, which can improve the conversion speed to video.

It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the present invention as defined by the following claims and their equivalents. Only. The scope of the present invention is defined by the appended claims rather than the detailed description and all changes or modifications derived from the meaning and scope of the claims and their equivalents are to be construed as being included within the scope of the present invention do.

The present invention can be applied to a service and a system for transmitting content including video to a user terminal from a content server.

Further, the present invention can be applied to a content server that performs a process of encoding multitrack video into SVC video in order to transmit content to a user terminal.

In addition, the present invention can be applied to a content media communication system including a content server for providing content including video and audio using a communication network, or a user terminal for receiving and displaying content.

100: video providing system 110: video encoding device
120: communication network 130: user terminal
210: communication unit 220: original video storage unit
230: Multitrack video encoder 240: Video alignment unit
250: Bitstream analysis unit 260: SVC encoding unit

Claims (4)

The method comprising: receiving a bitstream for video of a plurality of layers;
Determining an encoding mode and prediction information for a macroblock of a lower layer among the plurality of layers;
Performing prediction on a macroblock of the lower layer using the determined encoding mode and prediction information;
Determining an encoding mode and prediction information for a macroblock of an upper layer among the plurality of layers; And
And performing an inter-layer prediction on a video of an upper layer using an encoding mode and prediction information for a macroblock of the upper layer to reconstruct video, the scalable video decoding method comprising:
The encoding mode information for the macroblock of the lower layer indicates an inter prediction mode or an intra prediction mode,
Wherein the prediction information for the macroblock of the upper layer is determined using prediction information for the macroblock of the lower layer.
delete The method according to claim 1,
Wherein the encoding mode information for the macroblock of the upper layer indicates one of an inter prediction mode, an intra prediction mode, and an interlayer prediction mode.
The method according to claim 1,
Wherein the prediction information for the macroblock of the lower layer includes motion vector information and macroblock partition information.

Images