KR20050099488A - Method and apparatus for encoding and decoding of a video multimedia application format including both video and metadata - Google Patents

Abstract

The present invention relates to a method and system for encoding / decoding a video multimedia application file format for the integration of video and metadata. That is, the present invention provides for efficient indexing and retrieval in the process of constructing a new video multimedia application format (Video MAF) that combines multimedia data and metadata related to video data. Audio / image / motion feature information obtained from video data and contents is used as metadata of the video MAF file, and the video and metadata included in the MAF file by the user according to the structure of the video MAF. Allows you to browse the data effectively.

Description

METHOD AND APPARATUS FOR ENCODING AND DECODING OF A VIDEO MULTIMEDIA APPLICATION FORMAT INCLUDING BOTH VIDEO AND METADATA}

The present invention relates to a Video Multimedia Application Format (Video MAF), and more particularly, in the process of constructing a new video multimedia application format that combines multimedia content and metadata related to video data. Audio / image / motion feature information obtained from video data and contents is used as metadata of video MAF files for indexing and retrieval. A system and method for encoding / decoding video MAF files to enable efficient browsing of video and metadata contained in MAF files.

Moving Picture Experts Group (MPEG), an international standardization organization, has been standardizing MPEG-2, MPEG-4, MPEG-7 and MPEG-21 since the standardization of MPEG-1 in 1988. Technologies take the form of supporting specific application services by organizing profiles of standards into a set of elements required within a single standard specification.

However, such limited standardization has been a problem since it is difficult to satisfy various technical requirements of the industry with one standard technology. That is, as various multimedia standard technologies are developed, there is a need for combining different standard technologies and applying them to new multimedia application services.

To this end, MPEG-A (MPEG Application: ISO / ICE 23000) is conducting standardization of this need. The Multimedia Application Format (MAF) under way in MPEG-A combines the existing MPEG standards as well as non-MPEG standards such as JPEG to actively respond to the needs of the industry to increase the value of the standard. Doing. The main purpose of MAF standardization as above is to provide the industry with easy access to MPEG technology, so that it is easy to combine standard technologies that have already been verified without the need to create a new standard for application services required by the industry. .

Currently, the standardization of application file formats for music content is performed under the name of MPEG Music Player Application Format (ISO / IEC 23000-2), and the music MAF is in the final draft international standard (FDIS), and standardization is almost completed. This expands the functionality of MP3 players that were previously only played, allowing users to automatically categorize and play music by genre, show lyrics while listening to music, and browse album jacket photos related to music. There is a file format for users to get better music services. In particular, recently, MP3 players are becoming popular among consumers, such as being mounted in mobile phones, game machines (Sony's PSP), or portable multimedia players (PMPs). Therefore, it will be possible to commercialize an advanced music player using MAF soon.

In addition, standardization of application file formats for video, especially photo content, begins with the name MPEG Photo Player Application Format (ISO / IEC 23000-3), and the format standardization of photo MAF is in the working draft state. As with MP3 music, as the digital camera market continues to grow, photographic (generally JPEG) data acquired through digital cameras is growing rapidly. As the media (memory cards) that store photo data become smaller and more integrated, hundreds of photos can be stored on a single memory card, but in proportion to the increasing amount of photos, users experience difficulty managing photos. have. Accordingly, in order to solve this problem, photo MAF has attempted to standardize a new file format that bundles metadata related to photo contents together in a single file.

On the other hand, in addition to the standardization work for the music content and photographic content, the proposal for the standardization research is also started and the utilization of the video MAF such as personal video or broadcast video is expected to increase. Therefore, if you construct a new video multimedia application format that combines multimedia content and metadata related to video data, and allows users to effectively browse the video and metadata contained within the video MAF file, the multimedia content More effective use of is expected.

Accordingly, an object of the present invention is to solve the problems of existing video data and metadata management methods so that a user can effectively browse a video on a portable terminal. / Motion feature information and various feature information for effective indexing of photos are encoded as video MAF files using metadata, and multimedia contents related to video data so that users can browse and consume video MAF files effectively according to the above structure. And a method and system for encoding / decoding a video MAF file combining metadata.

In addition, another object of the present invention is to integrate the video data, images, subtitles and other multimedia contents including music files into one application file format called video MAF, the structure of the video data, video and By including audio / image / motion feature information from the content, summary information for effective indexing and retrieval of video as metadata, and metadata-based content utilization methodological tools, Even though it does not include the function of utilizing metadata, it is possible to effectively consume multimedia contents.

The present invention for achieving the above object, the step of the individual user acquires / inputs the personal video from the portable terminal such as PDA, digital camera, mobile phone, game machine or the service provider or content producer to obtain / input the broadcast video, Video processing and authoring to obtain feature information from the input video, generating metadata with the acquired feature information, encoding video MAF by combining the acquired video data and metadata, and transmitting the video MAF And decoding the transmitted video MAF, outputting the decoded video data, and outputting the decoded metadata.

In addition, the present invention is a key element of the invention, it defines the structure of the video MAF and the components of the metadata to be included in the video MAF, characterized in that it comprises a relationship and description structure of each component.

Hereinafter, with reference to the accompanying drawings will be described in detail the operation of the preferred embodiment according to the present invention.

1 is a block diagram showing a system configuration for effectively configuring, encoding, and decoding a video MAF according to an embodiment of the present invention.

Hereinafter, an operation in each block will be described in detail with reference to FIG. 1. First, the video acquisition / input part 1000 acquires or receives a personal video using a portable terminal device such as a PDA, a mobile phone, a digital camera, and a game machine. In addition, a service provider or a content producer acquires or receives a high quality broadcast video provided.

The video processing / authoring prat 2000 processes and authors feature information capable of representing input video data. That is, the video processing / authoring unit 2000 detects shots of video from input video, detects audio / image / motion descriptors of MPEG-7, preview images, and key-frames. ) Detect the images.

The video metadata creation part 3000 generates the feature information extracted by the video processing / authoring part 2000 as metadata. In this case, video metadata generation is generated in XML format based on the MPEG-7 MDS (Multimedia Description Scheme).

The video MAF encoding part 4000 encodes the video MAF by combining the input video and the generated metadata. As described above, the video MAF by the service provider or the content producer is transmitted to the user's portable terminal device in response to the user's request through the basic Q & A service system.

The video MAF decoding part 6000 decodes the transmitted video MAF into video data and metadata in a portable terminal. The video MAF player 7000 browses and displays decoded video data. The metadata parser 8000 interprets a story-based video representation and feature information for retrieving the video from the metadata.

According to the present invention, in the processing and authoring of the feature information representing the input video data, preview image extraction, shot detection and motion feature information and audio feature information extraction in units of shots, key frame extraction, key frames A process of extracting image feature information of a unit is provided.

FIG. 2 is a detailed structural diagram of the video processing / operation unit 2000 illustrated in FIG. 1. Hereinafter, the operation of each component of the video processing / storage unit will be described in detail with reference to FIG. 2.

First, the preview image selection part 2110 selects a preview image representing the input video data. In this case, the preview image includes poster images or advertisement images representing video data, and the format of the preview image includes, but is not limited to, JPEG, BMP, GIF, and the like.

The shot detection unit 2120 segments the input video data in units of shots, and acquires shot sections and key frames representing each shot as a result of the shot detection. At this time, the format of the key frame includes but is not limited to JPEG, BMP, GIF, and the like.

The motion feature extraction part 2130 obtains a motion feature from the shot periods acquired by the shot detector 2120 using MPEG-7 motion descriptors. In this case, a camera motion and a motion activity are extracted as a motion feature.

The audio feature extraction part 2140 acquires an audio feature using MPEG-7 low-level audio descriptors from the shot periods acquired by the shot detector 2120. Audio features include basic, basic spectral, basic signal, temporal timbral, spectral timbral, and spectral basis descriptors. It is characterized by.

The key-frame extraction part 2150 extracts a keyframe in the extracted shot section. An image feature extraction part 2160 extracts an image feature from the extracted keyframe. Image features include color structure, color layout, scalable color, dominant color, edge histogram, and homogeneous texture descriptors. .

The present invention is characterized by using MPEG-7 in describing metadata according to a standardized format and structure.

FIG. 3 illustrates components and structures of metadata generated by the video metadata generator 3000 of FIG. 1.

Referring to FIG. 3, the video MAF metadata 3100 includes a media information and generation descriptor 3110, which is a description of the media itself, and includes a media semantic description unit 3120. And a digital rights description unit 3130 for describing the rights.

The media information / creation description unit 3110 includes metadata related to generation by MPEG-7 MDS, and includes metadata describing a temporal structure of the media.

Table 1 below represents a description structure for expressing metadata required for the information generation description of the media in XML format.

<complexType name = "CreationInformationType"> <complexContent> <extension base = "mpeg7: DSType"> <sequence> <element name = "Creation" type = "mpeg7: CreationType" /> <complexType> <sequence> <element name = "Title" type = "mpeg7: TitleType" minOccurs = "1" maxOccurs = "unbounded" /> <element name = "PreviewImage" type = "mpeg7: TitleMediaType" minOccurs = "0" /> <element name = "Abstract" type = "mpeg7: TextAnnotationType" minOccurs = "0" maxOccurs = "unbounded" /> <element name = "Creator" type = "mpeg7: CreatorType" minOccurs = "0" maxOccurs = "unbounded" /> <element name = " CreationCoordinates "minOccurs =" 0 "maxOccurs =" unbounded "> <complexType> <sequence> <element name = "Location" type = "mpeg7: PlaceType" minOccurs = "0" /> <element name = "Date" type = "mpeg7: TimeType" minOccurs = "0" /> </ sequence> </ complexType> </ element> <element name = "CreationTool" type = "mpeg7: CreationToolType" minOccurs = "0" maxOccurs = "unbounded" /> <element name = "CopyrightString" type = "mpeg7: TextualType" minOccurs = "0" maxOccurs = "unbounded" /> </ sequence> </ complexType> <element name = "Classification" type = "mpeg7: ClassificationType" minOccurs = "0" /> <element name = "RelatedMaterial" type = "mpeg7: RelatedMaterialType" minOccurs = "0" maxOccurs = "unbounded" /> </ sequence> </ extension> </ complexContent> </ complexType>

The metadata representing the temporal structure of the media contains information that segments the temporal structure of the media, includes the MPEG-7 audio / motion features required for each segment, and extracts the keyframe image from each segment. And the MPEG-7 image feature to be extracted.

Table 2 below represents a description structure for expressing the temporal structure of the media in XML format.

<complexType name = "TemporalDecomposition"> <complexContent> <sequence> <element name = "AudioVisualSegment" type = "mpeg7: AudioVisualSegmentType"> <complexType> <sequence> <element name = "TextAnnotation" type = "mpeg7: TextAnnotationType" / > <element name = "MediaTime" type = "mpeg7: MediaTimeType" /> <element name = "MediaSourceDecomposition" type = "mpeg7: VideoSegmentMediaSourceDecompositionType"> <sequence> <element name = "VideoSegment" type = "mpeg7: VideoSegmentType" minOccurs = "0" maxOccurs = "unbounded" /> <element name = "MediaTime" type = "mpeg7: MediaTimeType" /> <element name = "Keyframe" type = "mpeg7: ImageLocatorType" /> <element name = "VisualDescriptor" type = "mpeg7: VisualDType" minOccurs = "0" maxOccurs = "unbounded" /> </ Keyframe> <element name = "VisualDescriptor" type = "mpeg7: VisualDType" minOc curs = "0" maxOccurs = "unbounded" /> </ VideoSegment> <element name = "AudioSegment" type = "mpeg7: AudioSegmentType" minOccurs = "0" maxOccurs = "unbounded"> <element name = "MediaTime" type = "mpeg7: MediaTimeType" /> <element name = "AudioDescriptor" type = "mpeg7: AudioDType" minOccurs = "0" maxOccurs = "unbounded" /> </ AudioSegment> <sequence> </ MediaSourceDecomposition> </ sequence> </ complexType> </ AudioVisualSegment> </ sequence> </ complexContent> </ complexType>

Media semantic description 3120 The media semantic description section may include a description of a semantic based feature of a media resource such as a main character or summary information of the media.

Table 3 below represents a description structure for expressing the meaning of the media resource in XML format.

<complexType name = "SemanticDescriptionType"> <complexContent> <extension base = "mpeg7: ContentAbstractionType"> <choice> <element name = "Semantics" type = "mpeg7: SemanticType" minOccurs = "1" maxOccurs = "unbounded" /> <element name = "ConceptCollection" type = "mpeg7: ConceptCollectionType" minOccurs = "1" maxOccurs = "unbounded" /> </ choice> </ extension> </ complexContent> </ complexType>

Digital right description 3130 is characterized by including a description of the access rights of the video encoded in the video MAF.

[Table 4] below represents the description structure of the access rights of the media in XML format.

<complexType name = "DigitalRightType"> <complexContent> <extension base = "mpeg7: AvailabilityType"> <sequence> <element name = "EditPermission" type = "boolean" use = "required" /> <element name = "CopyPermission" type = "boolean" use = "required" /> <element name = "ReleasePermission" type = "boolean" use = "required" /> <element name = "ViewPermission" use = "required" /> <choice> <element name = "PulicView" type = "boolean" /> <element name = "LimitedView" type = "boolean" /> <element name = "PrivateView" type = "boolean" /> <element name = "FreeView" type = " boolean "/> </ choice> </ ViewPermission> <sequence> </ extension> </ complexContent> </ complexType>

According to the present invention, in the step of describing a video MAF structure combining metadata and multimedia contents related to video, the MPEG-4 file format is used.

FIG. 4 illustrates components and structures of video MAF files generated from the video MAF encoding unit of FIG. 1.

Referring to FIG. 4, the move 5010 includes a meta box 5030 and a track box 5040, and the move includes multiple tracks.

The meta box 5030 includes the video MAF metadata 3100 described in FIG. 3. In addition, the metabox includes an output scheme of a video MAF encoded with a plurality of video data using metadata. Here, the output method includes, but is not limited to, a time-based output method, an event-based output method, a story-based output method, and the like of each video data.

The track box 5040 may include a media box (mdia) 5050 including decoding information of media such as a video stream, an audio stream, and a subtitle stream in a file format. The media box 5050 has a structure including decoding information such as a time stamp of a corresponding media.

In particular, FIG. 4 illustrates a case of a video MAF encoding a plurality of multimedia contents and metadata related to one video stream, and for one video stream 5070, N audios. And streams (audio stream) 5080 and N subtitle streams (5090). In this case, each of the audio streams and the subtitles is content representing respective languages. For example, for one movie content, there may be various languages such as Korean, English, Japanese, Chinese, and the like. Thus, one video stream may have a plurality of audio and subtitle streams.

In the present invention, in describing the video MAF structure, a plurality of video streams are included in the video MAF.

FIG. 5 illustrates a structure of components and elements of a video MAF including a plurality of video streams, and illustrates a case in which a video MAF for one video stream described in FIG. 4 is extended.

Referring to FIG. 5, the data sets 5110 and 5120 include multimedia data and metadata, respectively. That is, the video MAF is characterized by including a plurality of sets of video data. For example, there are various videos of shooting scenes of soccer games. Therefore, the metadata for each shot scene video, the preview image, and keyframe images may be combined, and the combined results may be generated as a video MAF. 5 illustrates a combined video MAF for A videos, where each video includes N languages and M keyframe images.

Meanwhile, in the above description of the present invention, specific embodiments have been described, but various modifications may be made without departing from the scope of the present invention. Therefore, the scope of the invention should be determined by the claims rather than by the described embodiments.

As described above, the present invention provides effective indexing in the process of constructing a new Video Multimedia Application Format (Video MAF) that combines multimedia data and metadata related to video data. And audio / image / motion feature information obtained from video data and contents for retrieval as metadata of the video MAF file, and included in the MAF file according to the structure of the video MAF. There is an advantage to enable effective browsing of the video and metadata.

1 is an overall system configuration according to an embodiment of the present invention,

2 is a detailed structural diagram of a video processing / authoring unit according to an embodiment of the present invention;

3 is a structural diagram of components and elements of metadata according to an embodiment of the present invention;

4 is a structural diagram of components and elements of a video MAF according to an embodiment of the present invention;

5 is a structural diagram of components and elements of a video MAF including a plurality of video streams according to an embodiment of the present invention.

Claims (28)

A method of encoding / decoding multimedia content and metadata associated with a video into a single integrated video multimedia application format. Acquiring / inputting a video terminal from a portable terminal such as a PDA, a digital camera, a mobile phone, a game machine, or acquiring / inputting video data from a service provider or a content producer; A video processing and authoring step of obtaining feature information from the input video; Generating metadata by using the acquired feature information; Encoding a video MAF by combining the obtained video data and metadata; Receiving and decoding the encoded video MAF; Outputting the decoded video data and metadata Video MAF file encoding / decoding method comprising a. The method of claim 1, The video processing and authoring step of obtaining feature information from the input video includes: Selecting the preview image, detecting the shot, extracting the motion feature by the shot unit, extracting the audio feature by the shot unit, extracting the keyframe by the shot unit, and keyframe A video MAF file encoding / decoding method comprising extracting image features on a per-unit basis. The method of claim 2, Motion feature information extracted in the process of extracting the motion feature information, A video MAF file encoding / decoding method comprising MPEG-7 camera motion and motion activity information. The method of claim 2, Audio feature information extracted in the process of extracting the audio feature information, Extracted using MPEG-7 low-level audio descriptors. Audio features include basic, basic spectral, basic signal, temporal timbral, A method for encoding / decoding video MAF files, comprising spectral timbral and spectral basis descriptors. The method of claim 2, In the process of extracting the image feature information, the extracted image feature information, Including MPEG-7 color structure, color layout, scalable color, dominant color, edge histogram, and homogeneous texture descriptors A video MAF file encoding / decoding method. The method of claim 1, Video MAF metadata in the step of generating metadata with the obtained feature information, A video MAF file encoding / decoding method comprising media information and generated description information, which is a description of the media itself, media semantic description information, and digital rights description information describing the digital rights of the media. The method of claim 6, In the media information and generation description information, A method for encoding / decoding a video MAF file, characterized by implementing a description structure in XML format for representing metadata required for generating and describing information on the media. The method of claim 6, In the media information and generation description information, A method of encoding / decoding a video MAF file, characterized in that a narrative structure for expressing a temporal structure of a media is expressed in an XML format. The method of claim 6, In the media semantic description information, A method for encoding / decoding a video MAF file, characterized by representing a description structure for expressing the meaning of a media resource in an XML format. The method of claim 6, In the digital rights description information, A method for encoding / decoding a video MAF file, characterized by representing the description structure of the media's access rights in XML format. The method of claim 1, The encoded video MAF data in the step of encoding the video MAF by combining the obtained video data and metadata, A video MAF file encoding / decoding method comprising a plurality of audio streams, a plurality of subtitle streams, a plurality of keyframes, and a plurality of preview images for one video data. The method of claim 2, In the encoding of the video MAF, the decoding information for the preview image and the keyframe is included in metadata in the metabox, and the decoding information for the media data such as video data, audio streams, and subtitle streams is included in the mediabox. A method for encoding / decoding video MAF files, characterized in that it is included in. The method of claim 2, In the encoding of the video MAF, video MAF file encoding / decoding method, characterized in that for encoding the video MAF in a plurality of forms, each metadata associated with the encoded video data. The method of claim 1, In the outputting of the decoded video data and metadata, the output and presentation method of the plurality of video data may be any one of a time order output method, an event unit output method, and a story based output method of the respective video data. Video MAF file encoding / decoding method, characterized in that implemented in the manner of. A system for encoding / decoding multimedia content and metadata associated with a video into a single integrated video multimedia application format. A video acquisition / input unit for acquiring / inputting a video terminal from a portable terminal such as a PDA, a digital camera, a mobile phone, or a game machine, or acquiring / inputting video data from a service provider or a content producer; A video processing / authoring unit for processing and authoring feature information capable of representing video data applied through the video acquisition / input unit; A video metadata generator for generating metadata using feature information of the input video data obtained through the video processing / authoring unit; A video MAF encoding unit for encoding a video MAF by combining the input video data and metadata; A video MAF decoding unit for decoding the encoded video MAF into video data and metadata;  A video MAF playback section for browsing and displaying the decoded video data from the video MAF decoding section; Metadata analysis unit for interpreting the representation of the story-based video and the feature information for searching the video from the metadata Video MAF file encoding / decoding system comprising a. The method of claim 15, The video processing / authoring unit, Acquiring feature information from the input video, selecting a preview image, detecting a shot, extracting a motion feature in units of shots, extracting an audio feature in units of shots, A video MAF file encoding / decoding system, comprising the steps of extracting keyframes in shot units and extracting image features in keyframe units. The method of claim 16, Motion feature information extracted in the process of extracting the motion feature information, A video MAF file encoding / decoding system comprising MPEG-7 camera motion and motion activity information. The method of claim 16, Audio feature information extracted in the process of extracting the audio feature information, Extracted using MPEG-7 low-level audio descriptors. Audio features include basic, basic spectral, basic signal, temporal timbral, A video MAF file encoding / decoding system comprising spectral timbral and spectral basis descriptors. The method of claim 16, In the process of extracting the image feature information, the extracted image feature information, Including MPEG-7 color structure, color layout, scalable color, dominant color, edge histogram, and homogeneous texture descriptors Featured video MAF file encoding / decoding system. The method of claim 15, The video MAF metadata generated by the video metadata generator is A video MAF file encoding / decoding system comprising media information and generated description information, which is a description of the media itself, media semantic description information, and digital rights description information describing the digital rights of the media. The method of claim 20, In the media information and generation description information, A video MAF file encoding / decoding system characterized by implementing a narrative structure in XML format for representing metadata required for generating and describing information on media. The method of claim 20, In the media information and generation description information, A video MAF file encoding / decoding system, characterized in that a narrative structure for representing a temporal structure of a media is represented in an XML format. The method of claim 20, In the media semantic description information, A video MAF file encoding / decoding system, characterized in that a narrative structure for expressing the meaning of a media resource is expressed in an XML format. The method of claim 20, In the digital rights description information, A video MAF file encoding / decoding system characterized in that it describes in XML format a description structure of access rights of media. The method of claim 15, Video MAF data encoded by the video MAF encoding unit, A video MAF file encoding / decoding system, comprising: a plurality of audio streams, a plurality of subtitle streams, a plurality of keyframes, and a plurality of preview images for one video data. The method of claim 16, In the encoding of the video MAF, the decoding information for the preview image and the keyframe is included in metadata in the metabox, and the decoding information for the media data such as video data, audio streams, and subtitle streams is included in the mediabox. Video MAF file encoding / decoding system, characterized in that it is included in. The method of claim 16, In the encoding of the video MAF, video MAF file encoding / decoding system, characterized in that for encoding the video MAF in a plurality of forms, each metadata associated with the encoded video data. The method of claim 15, In the outputting video data and metadata from the video MAF decoding unit, the output and presentation methods of the plurality of video data may include a time order output method, an event unit output method, and a story based output method. Video MAF file encoding / decoding system, characterized in that implemented in any one of the manner.