JP2006246008A - Video transcoding system, video acquisition device, transcoder device, and video transcoding method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a video transcoding system, a video acquisition device, a transcoder device, and a video transcoding method for performing transcoding corresponding to the scene of video contents without imposing any processing load in distributing the video contents. <P>SOLUTION: A video acquisition device 10 acquires video contents and the operation signal of a camera, and generates meta data in which a policy in transcoding the predetermined scene of the video contents is described. A video database device 20 generates meta data by analyzing the characteristics of the video contents stored in a video database 21. A video information distribution server device 30 distributes the meta data and video contents acquired by the video acquisition device 10 or the video database device 20. The transcoder device 40 analyzes a policy described in the distributed meta data, and performs transcoding of the video contents. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a video transcoding system, a video acquisition device, a transcoder device, and a video transcoding method for transcoding video content to be distributed via a communication network.

  In recent years, with the spread of the Internet, high-speed communication environments have been established in ordinary homes such as FTTH (Fiber To The Home) and ADSL (Asymmetric Digital Subscriber Line). In addition to the fixed communication field, third-generation mobile phones have become widespread not only in the field of mobile communication, but it has become possible to perform high-speed communication. In addition to the development of such a high-speed communication infrastructure environment, the advancement of digital signal compression technology such as image compression technology represented by JPEG (Joint Photographic Experts Group) and video compression technology represented by MPEG (Moving Picture Experts Group) It is now possible to distribute not only text data but also contents with large data capacity such as images, audio, and video through a communication network.

Against this background, demand for video streaming services that emphasizes real-time performance with IP (Internet Protocol) technology is also increasing. However, there are the following two problems for distributing video to an IP network.
1. 1. There are networks with various bandwidths on the Internet, and the optimal video bit rate varies depending on the network. The IP network is a Besteffort type network, and due to the nature of sharing the network among users, there is a possibility that network congestion may occur, resulting in packet dropping, packet delay, etc., and streaming video distribution Network quality of service (QoS) may not be maintained

In order to solve the above two problems, a technique for reducing the amount of information (transcoding) by flexibly converting the video bit rate according to the network conditions has been studied. There are roughly the following two methods for transcoding the video bit rate.
1. 1. Reduction of information on movement Reducing the amount of information related to image quality

  However, the conventional technology does not take into account video features (scenes with high motion, scenes with no motion, scenes with a dark overall image, scenes with a bright overall image, etc.). Most are transcoding techniques. In addition, although a method for transcoding the video bit rate in consideration of video features has been proposed, it is necessary to analyze video feature quantities such as motion vectors in real time. There is a problem that transcoding processing becomes heavy compared to the coding method.

In order to prevent the transcoding process from becoming heavy, a technique using metadata previously added to image data has also been proposed (see, for example, Patent Document 1). In Patent Document 1, metadata for emphasizing a subject among generated image data is generated.
JP 2003-250132 A (paragraphs 0017 and 0022)

However, even if the technique described in Patent Document 1 is used, when distributing image data in real time via a network, it is necessary to generate metadata by analyzing the acquired image data in real time. Therefore, even if the load of transcoding processing can be reduced using metadata, the processing load is still applied to the entire video content distribution processing.
Also, with the method based on image analysis, it is difficult to reflect the intention and policy of the video content creator (this scene emphasizes image quality, this scene emphasizes motion, etc.) when transcoding the video bit rate.
The present invention has been made in view of the above problems, and a video transcoding system capable of performing transcoding according to a scene of video content without applying a processing load at the time of video content distribution, A video acquisition device, a transcoder device, and a video transcoding method are provided.

  In order to solve the above problem, the invention described in claim 1 is directed to a video transcoding system including a video acquisition device, a video database device, a video information distribution server device, and a transcoder device. The apparatus transcodes a predetermined scene of the video content based on a database that stores the video content, a video analysis unit that analyzes characteristics of the video content stored in the database, and an analysis result by the video analysis unit Archive video metadata description means for generating metadata describing a policy at the time, and the video acquisition device includes video acquisition means for acquiring video content being shot, and the video input by a user Input information for obtaining a policy for a given scene of content And a live video metadata description unit that generates metadata describing a policy acquired by the input information acquisition unit, wherein the video information distribution server device is the video shooting device or the video database device. Distribution means for distributing the metadata and video content acquired from the above, the transcoder device analyzing metadata distributed from the video information distribution server device, and analysis by the metadata analysis means A video transcoding system comprising transcoding means for transcoding video content distributed from the video information distribution server device based on the result is provided.

  According to this configuration, for video content acquired by the video acquisition means, metadata is generated according to the policy input by the user, and for video content stored in the database, the video content is analyzed and metadata is generated. Therefore, transcoding according to the scene of the video content can be performed using the metadata without applying a processing load at the time of distributing the video content.

  According to a second aspect of the present invention, there is provided a video transcoding system for transcoding video content to be distributed via a communication network, characteristics of a predetermined scene of the video content stored in a database, information input by a user, Based on at least one of the above, archive video metadata description means for generating metadata describing a policy for transcoding the predetermined scene, and when acquiring the video content, the video content is associated with the predetermined scene, Live video metadata description means for generating metadata describing a policy input by a user, and video content based on the metadata generated by the live video metadata description means or the archive video metadata description means Transco To provide a video transcoding system characterized in that a transcoding means for loading.

  According to this configuration, the video transcoding system generates metadata according to the policy input by the user for the video content acquired by the video acquisition unit, and a predetermined scene for the video content stored in the database. Since metadata is generated based on at least one of features and information input by the user, transcoding is performed according to the scene of the video content using the metadata without applying a processing load at the time of video content distribution. It becomes possible.

  According to a third aspect of the present invention, there is provided a video acquisition device that acquires video content to be distributed via a communication network, a video acquisition unit that acquires video content shot by the shooting device, and the image acquisition device that is transmitted from the shooting device. An input information acquisition unit that acquires an operation signal generated by a user operation, and a meta that describes a policy for transcoding a predetermined scene of the video content based on the operation signal acquired by the input information acquisition unit. Provided is a video acquisition device comprising live video metadata description means for generating data.

  According to this configuration, the video acquisition device can generate metadata describing a policy for transcoding a predetermined scene of the video content by acquiring an operation signal generated by a user operation. There is no need to analyze and transcode video content in real time at the time of distribution, and it is possible to perform transcoding according to the scene of the video content using metadata without applying a processing load.

According to a fourth aspect of the present invention, in a transcoder device that transcodes video content to be distributed via a communication network, metadata for analyzing a policy described in metadata associated with a predetermined scene of the video content There is provided a transcoder apparatus comprising analysis means and transcoding means for transcoding a predetermined scene of the video content based on an analysis result by the metadata analysis means.
According to this configuration, since the transcoder device transcodes the video content based on the policy described in the metadata, it is not necessary to analyze and transcode the video content in real time, and processing at the time of video content distribution The load can be reduced.

  According to a fifth aspect of the present invention, in the video transcoding method for transcoding video content to be distributed via a communication network, the video acquisition device acquires the video content acquisition step, and the video acquisition device includes: An input information acquisition step of acquiring a policy for a predetermined scene of the video content input by a user, and a video in which the video acquisition device generates metadata describing the policy acquired in the input information acquisition step A metadata description step and a transcoder device comprising: a transcoding step of transcoding the video content based on a policy described in the metadata generated in the metadata description step. Providing an image transcoding method.

  According to this method, since the metadata describing the policy input by the user is generated and the video content is transcoded based on the generated metadata, a processing load is applied when distributing the video content. Rather, transcoding reflecting the user's intention can be performed according to the scene of the video content.

  According to the present invention, the video transcoding system generates metadata according to the policy input by the user for the video content acquired by the video acquisition means, and the predetermined scene for the video content stored in the database. Since metadata is generated based on at least one of features and information input by the user, transcoding is performed according to the scene of the video content using the metadata without applying a processing load at the time of video content distribution. It becomes possible.

Next, embodiments for carrying out the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing a configuration of a video transcoding system 1 according to an embodiment of the present invention. As shown in the figure, a video transcoding system 1 includes a video acquisition device 10 that acquires video content, a video database device 20 that stores video content, and a video information distribution server device 30 that distributes video content and metadata. And a transcoder device 40 that transcodes video content distributed to the receiving device 50. These devices are communicably connected via a network or the like. Hereinafter, the configuration of each device constituting the video transcoding system 1 will be described.

[Video acquisition device]
The video acquisition device 10 is a CPU (Central Processing Unit) that controls the entire video acquisition device 10, a communication interface that controls communication with the video information distribution server device 30, and a photographing device connected to the video acquisition device 10. A camera interface for receiving video content and an operation signal generated by a photographer's operation, and a storage device for storing software.

The camera function is not limited to the configuration in which the camera is connected to the video acquisition device 10, and the video acquisition device 10 itself may be equipped with a camera function. For example, the video acquisition device 10 may be a camera-equipped mobile phone.
Various data and programs are stored in the storage device of the video acquisition device 10. The various programs include a program for encoding video content and a program for generating metadata.

As the software resources and hardware resources included in the video acquisition device 10 operate in cooperation, the video acquisition device 10 includes the video acquisition unit 11, the input information acquisition unit 12, the live video metadata description unit illustrated in FIG. 13 and the encoding means 14.
The video acquisition unit 11 acquires video content shot by the camera.
The input information acquisition unit 12 acquires input information such as a policy input by the user when the video content is acquired by the video acquisition unit 11. Specifically, the input information acquisition unit 12 acquires information such as an operation signal transmitted from the camera and a policy input by a person involved in video creation to the video acquisition device 10.

  For example, a video shooter takes a video with a camera, and by pressing a predetermined button on the camera, he or she sets a policy (policy) such as focusing on movement in the scene currently being shot and focusing on images. Input as an operation signal. Also, a person involved in video creation grasps the scene characteristics of the video content shot by the camera and inputs information corresponding to the scene characteristics to the video acquisition apparatus 10. In this way, by describing the intention and policy of the video content creator such as image quality emphasis and motion emphasis in the metadata, it becomes possible to transcode the video content in accordance with the video content creator's intention.

The live video metadata description unit 13 generates metadata describing the policy acquired by the input information acquisition unit 12. FIG. 2 shows an example of a policy described in metadata. In the example shown in FIG. 2, the scene from 550 seconds to 1000 seconds in the video content is a motion-oriented scene because the movement is intense, and the scene from 5150 seconds to 12000 seconds is a scene in which the image quality is emphasized because there is little movement. It shows that there is.
The encoding unit 14 encodes the video content acquired by the video acquisition unit 11.

[Video database device]
Next, the configuration of the video database device 20 will be described. The video database device 20 is a database server, and includes a CPU that controls the entire video database device 20, a storage device that stores software, and a communication interface that controls transmission and reception of data via a communication network.
Various data and programs are stored in the storage device of the video database device 20. The various programs include a program for analyzing video content and a program for generating metadata. As the software resources and hardware resources included in the video database device 20 operate in cooperation, the video database device 20 includes the video database 21, the video analysis means 22, and the archive video metadata description means shown in FIG. 23 functions.

The video database 21 includes a storage device such as a hard disk, and is a database for accumulating and storing video content as archive video.
The video analysis means 22 analyzes feature quantities such as motion vectors of video content stored in the video database 21.
The archive video metadata description unit 23 generates metadata describing a policy used when transcoding video content based on the analysis result by the video analysis unit 22. For example, the frame rate is often more important than the image quality for scenes with a lot of movement, and the image quality is lowered for scenes without movement, and the frame rate is lowered because the image quality is more important than the frame rate. Can be considered. In this way, the archive video metadata description means 23 can perform transcoding based on the characteristics of the video content.

Since the archive video metadata description means 23 performs video analysis offline, it can also perform processing that requires time for analysis, and can generate metadata describing an effective policy with high accuracy. Specifically, not only is there a lot of movement or focus on image quality, but also what items are used for information reduction (shaping), such as video image quality, frame rate, Inra coding rate, etc., and which items are combined to reduce information It is possible to describe a detailed policy about such as.
Further, similarly to the video acquisition device 10, the archive video metadata description means 23 can generate metadata based on information input by the video content creator. It becomes possible to perform transcoding that reflects the intention of the content creator.

[Video information distribution server]
Next, the configuration of the video information distribution server device 30 will be described. The video information distribution server device 30 includes distribution means for distributing video content and metadata. The distribution means includes a streaming server 31 and a metadata distribution server 32.
The streaming server 31 includes streaming distribution means 311. The streaming distribution unit 311 converts the video content acquired from the video acquisition device 10 or the video database device 20 into a streaming format and distributes it to the transcoder device 40.
The metadata distribution server 32 includes metadata distribution means 321. The metadata distribution unit 321 distributes the metadata acquired from the video acquisition device 10 or the video database device 20 to the transcoder device 40.

[Transcoder device]
Next, the configuration of the transcoder device 40 will be described. The transcoder device 40 is a device that performs transcoding of video content based on streaming-format video content received from the streaming server 31 and metadata received from the metadata distribution server 32.
The transcoder device 40 includes a CPU, a storage device, and a communication interface, and includes a general computer hardware configuration. The storage device of the transcoder device 40 stores a program for analyzing metadata, a program for transcoding video content, and information indicating a policy for performing transcoding.

The metadata analysis unit 41 analyzes the policy described in the metadata received from the metadata distribution server 32.
The transcoding unit 42 transcodes a predetermined scene of the video content distributed from the streaming server 31 based on the analysis result by the metadata analysis unit 41. At this time, the transcoding unit 42 determines which item, such as video image quality, frame rate, Inra encoding, Inter encoding, etc., and which items are combined to reduce information for a predetermined scene. Etc. are performed as described in the metadata.

For example, when the metadata only describes “motion-oriented” and “image quality-oriented” and does not describe detailed information related to transcoding, the transcoding unit 42 determines that the own device 40 The video content is transcoded according to the policy stored in advance in the storage device. Examples of policies for transcoding include the following.
A) In the case of emphasizing motion ・ Reducing the amount of information related to image quality ・ Reducing the resolution of the image ・ Deleting the high-frequency component of the image (In the case of MPEG, after inverse DCT (discrete cosine transform), the quantization step is increased, and the DCT is performed again. Do)
・ Apply a lot of Inter coding B) When image quality is important ・ Reduce frame rate ・ Apply a lot of Intra coding

[Processing of video acquisition device]
Next, the flow of processing performed by the video acquisition device 10 will be described with reference to the flowchart shown in FIG.
First, the videographer operates the camera to give a shooting instruction. The camera starts shooting video and transmits video content to the video acquisition device 10.
The video acquisition device 10 acquires the video content transmitted from the camera by the video acquisition unit 11 (step S101), and encodes the video content acquired by the encoding unit 14.

In order to select whether the video scene currently being photographed is a scene where importance is placed on image quality or a scene where importance is placed on motion, the photographer selects a predetermined time loaded on the camera for a period of time when importance is placed on image quality or motion. I press the button. The camera transmits a camera operation signal generated by pressing the button to the video acquisition device 10.
The video acquisition device 10 receives the operation signal transmitted from the camera by the input information acquisition unit 12 (step S102).
Next, the video acquisition device 10 determines the policy, the video start time, and the end time based on the policy indicated by the received camera operation signal, and the reception start time and reception end time of the camera operation signal. Is generated (step S103).

[Transcoder processing]
Next, the flow of processing performed by the transcoder device 40 will be described with reference to the flowchart shown in FIG.
The transcoder device 40 receives video content and metadata associated with the video content (step S201).
The transcoder device 40 analyzes the metadata received in step S201 by the metadata analysis unit 41 (step S202).
The transcoder device 40 transcodes the metadata received in step S201 based on the analysis result of the metadata (step S203).

[Processing of video transcoding system]
Next, a flow of video content transcoding processing in the entire video transcoding system 1 will be described with reference to FIG.
In step 1, metadata describing a policy for transcoding video content is added to the video content.
In the case of live video, when a video is shot with a camera, a video shooter or a person involved in video creation inputs a policy in real time. For example, a video shooter inputs video as an operation signal for a camera while taking a video and placing importance on motion and image. The camera transmits the operation signal and each scene of the video content to the video acquisition device 10. The video acquisition device 10 acquires a camera operation signal by the input information acquisition unit 12 and acquires video content by the video acquisition unit 11. The video acquisition device 10 uses the live video metadata description unit 13 to generate metadata describing the policy indicated by the operation signal, and the video information distribution server device 30 combines the metadata and the video content encoded by the encoding unit 14. Send to.

  On the other hand, in the case of an archive video, the video feature is extracted from the video stored in the video database 21 using image analysis technology offline, and a policy that is estimated to be optimal transcoding is determined from the feature amount. Generate metadata describing the policy. For example, if the format of the video content is MPEG format, it is estimated from the amount of motion vectors that the scene is intensely moving, and metadata described as “active” as a policy is generated.

In addition, in the case of rich content such as a movie, the video content creator inputs a policy for transcoding such as placing importance on the image quality of this scene and placing importance on movement of the scene, to the video database device 820.
The video database device 20 stores the generated metadata and video content in the video database 21 in association with each other. The video database device 20 transmits metadata and video content to the video information distribution server device 30 when a predetermined trigger occurs, such as when a distribution request is received from the receiving device 50.

  In step 2, the streaming server 31 constituting the video information distribution server device 30 converts the video content acquired in step 1 into a streaming format and distributes the converted video content to the transcoder device 40. Further, the metadata distribution server 32 constituting the video information distribution server device 30 distributes the metadata attached (associated) to the video content in step 1 to the transcoder device 40.

  In step 3, the transcoder device 40 analyzes the policy described in the metadata by the metadata analysis means 41. The transcoder device 40 uses the transcoding means 42 to transcode the bit rate (video encoding amount) of the video content based on the analysis result. The transcoder device 40 distributes the video content after transcoding to the receiving device 50. The receiving device 50 decodes the distributed video content with a decoder.

As described above, the transcoder 40 refers to the metadata and transcodes the video content, so that there are the following two merits.
First, it is possible to perform transcoding according to the scene of the video content without imposing a processing load when distributing the video content.
For example, in a scene with motion, the amount of motion information can be reduced without reducing the amount of motion information, and in a scene without motion, the amount of information in motion can be reduced without reducing the amount of information on image quality. It becomes possible. Although it is possible to reduce the bit rate according to the scene in real time by applying image processing technology without using metadata in this processing, the transcoding processing becomes very heavy. In the method using metadata, it is possible to create metadata by applying image processing technology offline in advance for archive videos, or the video content creator can input a policy to be described in the metadata. is there. For real-time video (live video), the photographer can input the policy for reducing the bit rate on the spot using the operation signal of the camera as needed, thereby reducing the processing of the transcoder device 40. Is possible.

Second, transcoding that reflects the intention of the video content creator and the video creator is possible. Rather than automatically deciding information reduction items and other policies based on the scene characteristics of the video content, the video content creator and video creator would like to prioritize image quality for this scene and focus on motion for this scene. It is possible to reflect the intention of the video content creator or the video creator.
In this manner, transcoding according to the scene can be performed without applying a processing load at the time of video content distribution, and in addition, transcoding reflecting the intention of the video content creator or video photographer can be performed.

[Modification]
Although the embodiments of the present invention have been described above, the present invention is not limited to such embodiments, and various modifications can be made within the scope of the technical idea. As modifications, for example, the following can be considered.
(1) The device configuration of the video transcoding system 1 in the above-described embodiment is merely an example. The video transcoding system 1 describes metadata based on input information such as operation signals for live video, and describes metadata based on video analysis results or input information for archive video. The video transcoding system 1 may be composed of two devices or three devices as long as it has a function and a function of transcoding video content based on metadata. For example, the transcoder device 40 may directly acquire video content and metadata from the video acquisition device 10 or the video database device 20 without using the video information distribution server device 30 as a component of the video transcoding system 1.
Moreover, although the video transcoding system 1 in the above-described embodiment has been described as an apparatus that distributes both live video and archive video, it may be a system that distributes only one of live video and archive video. Good.

(2) In the above-described embodiment, the transcoder device 40 has been described as always transcoding the video content and then delivering it to the receiving device 50. However, the present invention is not limited to this, and the video content depends on the network conditions. The video content may be transcoded only when it is determined that the network QoS for streaming delivery cannot be maintained.
As a result, in video streaming distribution services, when there is a need to lower the bit rate of video content due to changes in network conditions, it is optimal based on the intention of the video content creator or video photographer or the feature quantity of the video content With this policy, video content can be transcoded.

(3) The example of the policy described in the above-described embodiment is merely an example, and for example, there may be a policy for reducing audio data included in the video content. Further, there may be a policy in which a scene in which the entire video is bright places more importance on image quality than a scene in which the entire video is dark.

It is a block diagram which shows the whole structure of the video transcoding system which concerns on embodiment of this invention. It is a figure which shows the example of description of the metadata which concerns on the embodiment. The flow of the process which the video acquisition device concerning the embodiment performs is shown. It is a flowchart which shows the flow of the process which the transcoder apparatus based on the embodiment performs.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Video transcoding system 10 Video acquisition apparatus 11 Video acquisition means 12 Input information acquisition means 13 Live video metadata description means 14 Encoding means 20 Video database apparatus 21 Video database 22 Video analysis means 23 Archive video metadata description means 30 Video information distribution Server device 31 Streaming server 311 Streaming delivery means 32 Metadata delivery server 321 Metadata delivery means 40 Transcoder device 41 Metadata analysis means 42 Transcoding means 50 Receiving device

Claims (5)

In a video transcoding system including a video acquisition device, a video database device, a video information distribution server device, and a transcoder device,
The video database device
A database for storing video content;
Video analysis means for analyzing the characteristics of the video content stored in the database;
Archive video metadata description means for generating metadata describing a policy for transcoding a predetermined scene of the video content based on an analysis result by the video analysis means,
The video acquisition device includes:
Video acquisition means for acquiring video content being shot;
Input information acquisition means for acquiring a policy for a predetermined scene of the video content input by a user;
Live video metadata description means for generating metadata describing the policy acquired by the input information acquisition means,
The video information distribution server device includes:
Distribution means for distributing metadata and video content acquired from the video imaging device or the video database device;
The transcoder device is
Metadata analysis means for analyzing metadata distributed from the video information distribution server device;
A video transcoding system comprising: transcoding means for transcoding video content distributed from the video information distribution server device based on an analysis result by the metadata analysis means. In a video transcoding system for transcoding video content to be distributed via a communication network,
An archive video that generates metadata describing a policy for transcoding the predetermined scene based on at least one of characteristics of the predetermined scene of the video content stored in the database and information input by the user Metadata description means;
Live video metadata description means for generating metadata describing a policy input by a user in association with a predetermined scene of the video content at the time of video content acquisition;
A video transcoding system comprising: transcoding means for transcoding video content based on metadata generated by the live video metadata description means or the archive video metadata description means. In a video acquisition device that acquires video content to be distributed via a communication network,
Video acquisition means for acquiring video content shot by the shooting device;
Input information acquisition means for acquiring an operation signal by a user's operation transmitted from the imaging device;
Live video metadata description means for generating metadata describing a policy for transcoding a predetermined scene of the video content based on the operation signal acquired by the input information acquisition means. A video acquisition device. In a transcoder device that transcodes video content to be distributed via a communication network,
Metadata analysis means for analyzing a policy described in metadata associated with a predetermined scene of video content;
Transcoding apparatus, comprising: transcoding means for transcoding a predetermined scene of the video content based on an analysis result by the metadata analysis means. In a video transcoding method for transcoding video content to be distributed via a communication network,
A video content acquisition step in which the video acquisition device acquires the video content;
An input information acquisition step in which the video acquisition device acquires a policy for a predetermined scene of the video content input by a user;
A video metadata description step in which the video acquisition device generates metadata in which the policy acquired in the input information acquisition step is described;
A transcoding device comprising: a transcoding step of transcoding the video content based on a policy described in the metadata generated in the metadata description step.

Images