KR20120068285A - Diversification transcoding method for video file - Google Patents

Abstract

PURPOSE: A distributed transcoding method of a video file is provided to reduce load for video transcoding by processing transcoding of a video file. CONSTITUTION: A transcoding server computer(100) transmits fraction data to sub computers(1021-102N). Each sub-computer transcodes the fraction data. Each sub computer offers the transcoded fraction data. The trandcoding server computer receives the trandcoded fraction data. The trandcoding server computer generates the trandcoded video file.

Description

Distributed transcoding method for video file

The present invention relates to a video file processing technique, and more particularly, to a distributed transcoding method of a video file.

A video communication system that provides a video service to a subscriber's terminal through a network serves various types of video according to a network environment, a terminal standard, and a service type.

Recently, as the demand for the convergence and interworking of services is increasing, such as using various video services in one terminal or designing a service that accommodates different network environments, video services are provided between subscribers of these various network / terminal environments. In order to provide smoothly, the situation is being prepared. What is needed to perform this interworking work is 'transcoding', which is a processing task that transforms the format and expression of the video file.

With the development of networks and personal terminals, various types of video communication services such as video call, video conferencing, video on demand, video chat, and multimedia messaging are being developed. The service has different characteristics according to the base environment such as the network or the subscriber's terminal.

That is, depending on whether the network bandwidth is sufficiently provided, or depending on the information processing capability of the terminal or the like, a video having a sufficient picture size or a frame or a video having a limited picture quality is provided.

When interworking between such video communication services is required, there is a case that a personal computer-based high quality video service must be delivered to a wireless terminal subscriber. In the interim process, the original video format is accepted by the final subscriber terminal. Transcoding is required to deform to the extent possible.

However, as many video communication services are developed and video formats provided by these services are diversified, different video encoding services need to be transcoded in consideration of network and terminal characteristics in order to interoperate with other video communication services. . From basic transcoding to convert the size of a video, to converting the format of a video, extracting summary information from the contents of a video file, extracting only some media, or adding unstructured compressed objects such as logos or commercial information. Transcoding may be required.

Therefore, it is not easy for one video communication service to perform many transcoding functions that can accommodate various interworking conditions. In addition, deploying these transcoding functions individually in multiple video communications services results in a significant overhead load.

Therefore, it is urgently needed to develop a technology for effectively performing transcoding required in various video communication services such as video call, video on demand, video conferencing and multimedia messaging.

It is an object of the present invention to provide a distributed transcoding method of a video file that can improve transcoding speed while reducing the load for video transcoding by distributing transcoding for a video file.

In the distributed transcoding method of a video file according to the present invention for achieving the above object, the transcoding server computer divides the audio data and the video data of the video file into a plurality of pieces: a part of the plurality of pieces Combining to construct fragment data; Transmitting the fragment data to a plurality of subcomputers; And each of the plurality of sub-computers transcodes the fragment data to receive the transcoded fragment data and combine them to generate a transcoded video file.

The present invention described above has the effect of distributing the transcoding for the video file to reduce the load for transcoding the video file and at the same time improve the transcoding speed.

1 is a block diagram of a transcoding system for a video file according to a preferred embodiment of the present invention.
2 is a flow chart of a transcoding method for a video file according to a preferred embodiment of the present invention.
3 is a diagram showing an example of a fragment configuration for a video file according to a preferred embodiment of the present invention.
4 and 5 are diagrams showing an example of a fragment data configuration for a moving picture file according to a preferred embodiment of the present invention.

The present invention distributes the transcoding of the video file, thereby reducing the load for transcoding the video file and at the same time improving the transcoding speed.

<Configuration of Transcoding System for Video Files>

The configuration of a transcoding system for a video file according to the present invention will be described with reference to FIG. 1.

The transcoding system for the video file is composed of a transcoding server computer 100 and a plurality of subcomputers 1021 to 102N.

The transcoding server computer 100 and the plurality of subcomputers 1021 to 102N are connected through a network to transmit and receive data, and distributed transcode video files through the plurality of subcomputers 1021 to 102N.

For the transcoding, the transcoding server computer 100 generates audio and video fragments for a video file, bundles the audio and video fragments together to generate fragment data, and transmits the fragment data to the plurality of subcomputers ( 1021 to 102N) to request transcoding, and receive transcoded fragment data from the plurality of sub-computers 1021 to 102N to combine them to form a new video file.

The plurality of sub-computers 1021 to 102N receive fragment data composed of audio and video pieces provided by the transcoding server computer 100 to perform transcoding, and transmit the result to the transcoding server computer 100. Return). In particular, since the size, options, and key frames of the fragment header change according to transcoding by the plurality of sub-computers 1021 to 102N, the size of the entire fragment is changed and extra data, which is data necessary for decoding the fragment header, is added. ) Is changed. Accordingly, the plurality of subcomputers 1021 to 102N update the header by reflecting the above-described changes in the header. The part of the updated fragment header is a frame flag (DWORD dwKeyFrame), the byte offset after the header (DWORD dwOffset), the byte size of the frame (DWORD dwSize), and the fragment header part includes an extra data length. Unsigned int length, BITMAPINFOHEADER (BITMAPINFOHEADER info) for video, WAVEFORMATEX (WAVEFORMATEX) for audio, and extra data (unsigned char * extra) for decoding.

<Distributed transcoding method of video file>

A transcoding method for a video file according to a preferred embodiment of the present invention will be described with reference to FIG.

The transcoding server computer 100 generates the video fragment in consideration of the audio fragment and the key frame in the stored spatial order (step 200).

The video fragment will be described with reference to FIG. 3. The video fragment may consist of a first or second video fragment.

The first video fragment includes a header, a first key frame, a second key frame as an additional key frame, and a plurality of frames. The plurality of frames refer to a first key frame or a second key frame.

The second video fragment includes a header, a second key frame, a third key frame, a fourth key frame which is an additional key frame, and a plurality of frames. The plurality of frames refer to the second key frame or the third key frame or the fourth key frame.

The header consists of a fragment header byte size (unsigned long header_size), a fragment number (unsigned long number), an unsigned long entry_number, and a frame index entry array (INDEX_ENTRY aIndex []).

The index entry for the frame consists of a key frame flag (DWORD dwKeyFrame), a byte offset after the header (DWORD dwOffset), and a byte size (DWORD dwSize) of the frame.

The video fragment according to the present invention has the following features. First, the frames in the fragment follow the order of space stored in the content, starting with the key frame and ending with the additional key frame. Keyframes may be included between the keyframes at both ends, and there may be no additional keyframes for the last piece.

The audio fragment is composed of an audio frame configured to correspond to the frames of the video fragment in the frame in the fragment according to the spatial order stored in the video file.

After generating the video and audio fragments for the video file as described above, the transcoding server computer 100 determines the number of fragments to be delivered in consideration of the specifications for each sub-computer and based on this. Construct fragment data. Here, the specification information of the sub computer may be registered and stored in the transcoding server computer 100 in advance, and the number of pieces corresponding to the specification may be predetermined.

As shown in FIG. 4, the fragment data may include a header and first to fifth pieces, a header and sixth to eighth pieces, or a header and ninth to twelfth pieces. That is, the fragment data is formed by combining different numbers of pieces according to the specifications of the subcomputer.

As shown in FIG. 5, the header of the fragment data includes fragment header byte size (unsigned long header_size), fragment number (unsigned long number), and audio and video identification information (unsigned long type; 0: audio, 1: video). , Unsigned long particle_number, the media type to be transcoded (FORCC target), the byte length including the extra data length, BITMAPINFOHEADER (BITMAPINFOHEADER info) for video, In the case of audio, it includes WAVEFORMATEX (WAVEFORMATEX) and extra data (unsigned char * extra), which is data necessary for decoding.

The WAVEFORMATEX provided in the fragment data header follows the WAVEFORMATEX structure format defined by Microsoft. The BITMAPINFOHEADER follows the format of a BITMAPINFOHEADER structure defined by Microsoft.

The transcoding server computer 100 transmits each piece of fragment data composed of the audio and video pieces to a plurality of sub-computers 1021 to 102N (step 202). The plurality of subcomputers 1021 to 102N transcode each piece of fragment data consisting of the video and audio pieces (step 204).

Subsequently, the plurality of sub-computers 1021 to 102N transcode fragment data composed of the transcoded file fragments according to the media type included in the fragment data and then transmit the fragment data to the transcoding server computer 100 (step 206). ). In this case, when the transcoding is performed, the size and offset of the fragment header frame and whether or not the key frame is changed, the overall size of the fragment is changed, and an extra, which is data necessary for decoding the fragment header, is changed. . Accordingly, the plurality of subcomputers 1021 to 102N reflect these changes in the received headers before delivering them to the transcoding server computer 100. These changes include the byte length (including the frame flag of the fragment header (DWORD dwKeyFrame), the byte offset after the header (DWORD dwOffset), the byte size of the frame (DWORD dwSize), and the extra data length of the fragment header. unsigned int length), BITMAPINFOHEADER (BITMAPINFOHEADER info) for video, WAVEFORMATEX (WAVEFORMATEX) for audio, and extra data (unsigned char * extra) for data.

The transcoding server computer 100 that receives the transcoded fragment data from the plurality of sub-computers 1021 to 102N combines the transcoded fragment data according to the fragment number to generate a transcoded video file (208). step).

100: transcoding server computer
1021 ~ 102N: Multiple Sub Computers

Claims (7)

In the distributed transcoding method of a video file,
The transcoding server computer divides the audio data and the video data of the video file into a plurality of pieces:
Combining some of the plurality of pieces to form fragment data;
Transmitting the fragment data to a plurality of subcomputers;
After each of the plurality of sub-computers have transcoded the fragment data, and if the transcoded fragment data is provided, receiving and combining the transcoded fragment data to generate a transcoded video file;
Distributed transcoding method of a video file, characterized in that it comprises a. The method of claim 1,
The transcoding server computer,
For each of the plurality of subcomputers, fragment data is generated in the number of pieces according to the specification of the subcomputer,
The specification information of the sub-computer and the number of pieces corresponding to the specification are stored in advance. The method of claim 1,
The fragment consists of a header and audio data or video data,
The header contains an unsigned long header_size, an unsigned long number, an unsigned long size, an unsigned long entry_number, and an index entry array for the frame. INDEX_ENTRY aIndex []),
The index entry for the frame includes a key frame flag (DWORD dwKeyFrame), a byte offset after the header (DWORD dwOffset), and a byte size (DWORD dwSize) of the frame. The method of claim 3,
The piece is
Or a video data and header consisting of a start key frame and a last additional key frame, and a plurality of frames existing between the key frame and the additional key frame, or
Or a video data and a header consisting of a start key frame and a last additional key frame, a plurality of frames existing between the key frame and the additional key frame, and a key frame positioned between the plurality of frames,
A distributed transcoding method of a video file, comprising a header and audio data interleaved with frames of video data included in a video fragment. The method of claim 1,
The fragment data consists of the header and fragments,
The header includes fragment header byte size (unsigned long header_size), fragment number (unsigned long number), audio and video identification (unsigned long type; 0: audio, 1: video), fragment number (unsigned long particle_number), , Unsigned int length including the media type (FORCC target) to change to transcoding, extra data length, BITMAPINFOHEADER (BITMAPINFOHEADER info) for video, WAVEFORMATEX (WAVEFORMATEX) for audio, decoding A method of distributed transcoding of a video file, comprising extra data (unsigned char * extra), which is required data. The method of claim 1,
The plurality of sub-computers,
After updating the header of the transcoded fragment data and the header of the fragments belonging to the fragment according to the frame size according to the transcoding or the change information on the key frame or not, the fragments belonging to the fragment are updated. And the transcoded fragment data to be transmitted to the transcoding server computer. The method of claim 6,
The portion of the fragment header to be updated is a frame flag (DWORD dwKeyFrame), a byte offset after the header (DWORD dwOffset), and a byte size (DWORD dwSize) of the frame,
The portion of the fragment header being updated includes an unsigned int length including extra data length, BITMAPINFOHEADER (BITMAPINFOHEADER info) for video, WAVEFORMATEX (WAVEFORMATEX) for audio, and extra data that is necessary for decoding. (extra) A distributed transcoding method of a video file characterized by being unsigned char * extra.

Images