TWI513287B - Automatic media editing apparatus, editing method, broadcasting method and system for broadcasting the same - Google Patents

Description

Media automatic editing device, media automatic editing method, media broadcasting method and broadcasting system thereof

The disclosure provides a media automatic editing device, a media automatic editing method, a media broadcasting method and a broadcasting system thereof, and particularly, a device for reproducing a video text segment captured by different regions according to a media feature by using a editing or marking method and Related methods.

The way to watch a program on a TV is usually to transmit the video signal to the home through cable TV or wireless TV. The wireless mode mainly uses the TV station to directly send the signal, and the antenna is directly viewed by the antenna installed in the home; the other way is mainly by cable. Channel operators transmit cable channel signals through cables that are connected to each user's home.

The Internet prevails and the user gradually has more network bandwidth. Beginning with content providers to set up Internet TV channels through the Internet, users can see TV programs that are instantly transmitted through the Internet. Watch the TV shows you want, anytime, anywhere with video on demand (VOD).

In the modern world where there is a large amount of media information, the amount of media delivered by the above-mentioned network method is huge, and viewers often need to spend a lot of time to search, integrate and receive the required content, if necessary, to summarize, splicing and Reuse can only be done manually, from content capture, editing, grouping, grouping, and re-creating new content. Each step of the process requires a lot of manpower and time to complete.

The disclosure proposes a media automatic editing device, a media automatic editing method, a media broadcasting method and a broadcasting system thereof, which are formed by a media automatic editing technology, according to a result of media feature comparison, or marking a specific source of media, after being re-formed. New multimedia clips to provide users. In particular, multimedia files that can be obtained from different regions (such as countries) can quickly and efficiently obtain the required multimedia content.

According to the embodiment described in the disclosure, the media automatic editing device has a feature database for recording various media features, wherein the feature values include system settings, and the update database can be learned by the user's preferences. The feature values recorded in the feature database include image features, sound features, and text features.

The media automatic editing device includes a content capturing module that can receive one or more multimedia segments, and then uses the content classification module to import feature values of the feature database and perform feature comparison to classify the received multimedia segments. The device has a memory module that temporarily stores multimedia clips.

After the feature comparison, the content re-modulation module in the device obtains one or more multimedia segments according to the request of a certain client, and the corresponding feature is requested to be one of the image features, the sound features and the text features, or a combination thereof. . After the re-production, a new multimedia segment is formed, and then the content-casting module broadcasts the corresponding new multimedia segment to the requesting client.

According to an embodiment of the automatic media editing method of the disclosure, the steps include first capturing one or more multimedia segments, and then importing one or more feature values. Similarly, the feature values include image features, sound features, and text features. The method continues to perform media feature analysis, analyzes the characteristics of the multimedia segment, and then compares one or more multimedia segments that meet the characteristics of the request according to the request of the user terminal, and temporarily stores the data into a memory, and the temporary storage of the memory includes Multimedia clips that were originally received or filtered. Then, one or more multimedia segments that meet the request are formed into a new multimedia segment. After the feature comparisons result in one or more multimedia segments, the new multimedia segments will be re-created in a clip or markup manner.

The disclosure further proposes a media broadcast method, in particular, broadcasting the multimedia content reproduced by the feature comparison. According to an embodiment, a connection between a main server and a plurality of regional servers in different regions is first established, and each regional server captures multimedia content of the region in which it is located, classifies and marks each segment, and then the main server Based on the set characteristics, the primary server polls the regional servers for requests generated by the client to obtain segments or tags that conform to the requested multimedia content. After that, the main server re-forms a new multimedia content and broadcasts the multimedia content to the client.

The media broadcast system embodiment includes a main server for providing user management functions, a user login interface, a program menu, collecting popular keyword functions, and receiving a request from a client, and a plurality of regional servers.

Multiple regional servers can be set up in different regions, such as different countries. Each region's servers capture multimedia content in the region, such as recording TV programs or network exchange packets in the region, which can be analyzed. According to the characteristics of the multimedia content classification and marking, the regional server can exchange the records of classification and marking.

The main server can obtain the records of the classification and marking of the plurality of regional servers, so that the local server where the corresponding multimedia content is located can be compared according to the request of the user, and then the user terminal and the corresponding regional server are connected.

The regional server of one embodiment may include a media automatic editing device including a feature database for recording feature values, a content capture module for receiving multimedia content, and a content classification module for classifying multimedia content after comparing features. The group, the memory module, and the re-made feature matching segment form a new multimedia content content re-modulation module, and then the corresponding multimedia content is broadcasted to the user end through the content broadcast module.

The disclosure proposes a media automatic editing device, a media automatic editing method, a media broadcasting method and a broadcasting system thereof, and an automated solution for the complete process of content editing, summarization and re-production, which can automatically capture and convert for different content sources. Editing the desired content in a file entity or virtual mode and removing unwanted segments, automatically sorting and grouping, and then recomposing the combined segments into new content. Therefore, the content collection and producers can organize and reuse the huge content at the lowest labor cost and provide it to the audience. Viewers don't need to look for content from a huge source, but can browse through all the content they need in a short time with the essence and abstract.

The media automatic editing technique described in the disclosure may refer to the schematic diagram shown in FIG. 1. The device performing the automatic editing of the media obtains the original multimedia segment 101 from various media sources, and filters the feature conditions 103 of the image feature, the sound feature or/and the character feature, and clips the segment that meets the feature condition by using the editing method, or Is a mark (such as using the time stamp of the clip) that matches the feature alignment, is represented by a clip or mark segment 105 in the figure, and then forms a new multimedia clip 107, which is the last broadcast to each User-side movie. In particular, the media source can be of various types, including multimedia content transmitted by using a network, television programs transmitted by using a video network, etc. The present invention utilizes the technology of automatic media editing to continuously capture new multimedia segments and automatically Filtering and re-creating into new multimedia clips to provide continuous multimedia content on the client side.

For the device for automatic media editing, reference may be made to the block diagram of the embodiment of the media automatic editing device shown in FIG. 2.

The media automatic editing device 20 shown in the figure includes a multimedia input for inputting multimedia content and a multimedia output output after being edited. According to an embodiment, the media automatic editing device 20 can form corresponding multimedia content according to the input of the user terminal, as shown in the figure. Show the demand input on the right.

The content capturing module 201 in the device 20 is configured to obtain multimedia segments received by the external, including video, audio and text, or a combination thereof. The source of the multimedia segment includes at least an image, a video, a voice file, a TV channel, and a network string. Stream and camera. The memory module 207 is used to store these multimedia segments or for media signals in the temporary processing. The content classification module 203 then obtains the multimedia segments received by the content capture module 201, and first obtains the features of the multimedia segments through analysis, and then according to the various feature values obtained from the feature database 205, the content classification module 203 then compares Classify multimedia clips and feature values.

The feature database 205 records one or more feature values, and the media automatic editing device 20 processes various multimedia content according to the feature values. The feature values mainly include image features, sound features, and character features. The feature values recorded in the feature database 205 can be updated, and the feature values used to obtain the multimedia content can be built in the feature database 205, and the feature description generated by the client can be collected. The request generated by the client may include one of an image feature, a sound feature, and a text feature, or a combination thereof.

According to an embodiment, the image feature may be an image color distribution of each frame in the media segment. When multiple frames in the multimedia image conform to a specific image color distribution, the related segment may be classified into a content, or even related. Part of the mark or clip it out. For example, in a football game, a certain proportion of the picture is grass green. If you want to exclude the segments that enter the advertisement afterwards, you can delete the consecutive frames of grass green that are less than a threshold. Will leave the content of the football match. In this example, the feature value in the multimedia image may further include a time information, and the multimedia segment to be obtained may be played during a specific program time period, so the time parameter is a parameter that can provide more accurate content to obtain the desired content.

In still another embodiment, the sound characteristics may be sound waveforms, amplitudes, and frequency changes. According to samples of sound waveforms and amplitudes of specific content, multimedia content conforming to the samples may be obtained by analysis. For example, music programs usually have a range of volume and frequency changes. After entering an advertisement, they usually have higher audio, and the frequency changes are less irregular. Therefore, the sound waveform, amplitude and frequency can be used as a basis for effectively screening multimedia segments. Similarly, such feature values will be matched with time information to obtain content of interest to the client in an accurate time.

According to another embodiment, the feature value may be a character feature. If the multimedia segment has a caption or a caption, the keyword may be used to obtain a multimedia segment that conforms to the feature. It is also possible to use the feature value of the time information as a parameter for screening.

After the multimedia segments are classified by the matching feature values, the content re-manufacturing module 209 in the media automatic editing device 20 obtains the request through the input module 211, that is, corresponding to the request to obtain one or more multimedia corresponding to the features requested by the end users. The fragment is reconstructed to form a new multimedia segment.

Then, the device 20 content broadcast module 213 broadcasts the corresponding requested new multimedia segment to the requesting client.

According to one embodiment of the media automatic editing device 20 described in the disclosure, the media automatic editing device 20 can be connected to a television playing device, such as a set-top box, a video conversion device, etc., and the content capturing module 201 of the device 20 For example, an image capturing device is connected to a related device for television broadcasting, and automatically captures a digital image file by means of a software or a hardware circuit, and converts the image into a desired format.

In the case of performing automatic content editing and categorization when the multimedia clip features are subsequently compared, the techniques cited therein include the use of audio discrimination scenes disclosed in the Republic of China Patent No. I319179 (Application No.: 095133800; Application Date: 2006/09/13). The algorithm, or the Republic of China Patent Publication No. 201018198, discloses a film subtitle identification method (application number: 097140019; application date: 2008/10/17). After the feature recognition and editing, each program content is divided into several segments or actually stored as several audio and video files, and the advertising portion can be simultaneously removed. Each of the above segments can be associated with the program list, subtitles and voice content recognition when generated, so that classification and grouping can be automatically performed when storing.

Finally, in the program summary re-creation, according to the feature value, classification, date and time parameters in the feature database 205, and the attribute data of the user end, a plurality of related segments are automatically brought into the slice-containing end by the sequence arrangement. The library, or layout into a multi-split screen, produces new video content, which is then passed to the viewing client.

[One of the media automatic editing embodiments] The embodiment of the media automatic editing method applying the above-described media automatic editing device can refer to the flowchart described in FIG.

At the beginning of the process, in step S301, the media automatic editing device obtains one or more multimedia segments through the content capturing module, and then imports the features generated by the feature database or integrated according to the user attributes and the search keywords. The value (step S303), the imported feature values such as shadow (31), sound (32), text (33), or a combination thereof.

Performing media feature analysis by means of software or hardware implemented in the media automatic editing device (step S305), obtaining features in each multimedia segment, such as color distribution and brightness of each frame in the movie, in the sound Waveform, amplitude and frequency changes, or keyword characteristics of the text.

Then, in step S307, the requirements generated by the UEs are compared, and the corresponding feature values may be recorded in the request message of the UE, and the one or more multimedia segments that meet the features described in the request are generated after the comparison. .

Then, the compared multimedia clips are temporarily stored in the memory (step S309) to provide the clips for the desired clips (step S311), and the clips are temporarily stored in the memory (step S313), and then reproduced. Another multimedia segment is formed (step S315), and a correction means (step S319) can be introduced during the editing process to check whether the user's needs are met and then broadcast (step S317).

The clipped multimedia clip may be a media content that excludes unnecessary content and corresponds to content required by the client. For example, the request made by the client is a football match, after the image (such as color), sound (such as the waveform of the broadcast, amplitude and frequency changes), text features (subtitles of the related broadcast content) and time (posting time) are compared. The clips that match the feature are clipped out to obtain a plurality of clips related to the football, and then combined to form a new multimedia clip.

According to the media automatic editing method proposed by the disclosure, after the feature comparison, the corresponding multimedia segment can be edited as in the above-mentioned editing mode of FIG. 3, and one or more multimedia segments conforming to the request are extracted and combined. For new multimedia clips.

According to another embodiment, instead of clipping the segments that match the feature, one or more multimedia segments that meet the request may be marked by a tagging method, and the timestamps are recorded, and the segments of the tags are to be recorded later. Combine to form new multimedia clips.

[Second Media Automatic Clipping Embodiment 2] Referring to the flowchart of the second embodiment of the media automatic editing method shown in FIG.

Beginning in step S401, the multimedia content is revealed by the media automatic editing device, and then the feature value including the shadow (41), the sound (42) or the text (43), or a combination thereof is imported (step S403), including The feature value of the feature database also includes the feature information collected by each client attribute and usage.

Then, the media feature analysis is performed (step S405), and the captured feature of the multimedia segment is compared with the imported feature value to obtain a multimedia segment that meets the requirements issued by the specific client (step S407).

In this example, instead of capturing the desired segment, the multimedia segment matching the feature is marked (step S409), such as the time stamp of the matching segment, and is recorded in the memory, and then the same type of marker segment is combined. When it is up (step S411), the corresponding multimedia segments are combined and temporarily stored in the memory (step S413). The content can also be confirmed by the correction program (step S417). Finally, the combined new multimedia clip is broadcast to the requesting client (step S415).

According to an embodiment of the present disclosure, the media automatic editing apparatus and method described above are particularly applicable to a media distribution system as shown in FIG. The embodiment shown in FIG. 5 describes a broadcast system including a main server 503, a plurality of regional servers (51, 52, 53) and a client device 501, each of which can be interconnected by an internet 50.

According to the embodiment shown in FIG. 5, there are a plurality of regional servers installed in various places, such as the first server 51 disposed in the area 511, the second server 52 disposed in the area 512, and the area server. The third server 53 of the third 513, the main server 503 and the regional servers (51, 52, 53) of each region form a cloud-based media broadcast system. A plurality of regional servers are connected to the main server 503 via the Internet 50, and can exchange information and support each other. The main server 503 can provide user management functions, a user login interface, a program menu, collect popular keyword functions, and receive a request from a client.

The above-mentioned local area server establishes a cloud-operated multimedia system, and provides a client-side multimedia download service by the main server. The master server and a plurality of regional servers are a master-slave architecture system, and the role of the master server can be performed by one of a plurality of regional servers, which is responsible for multimedia file attributes and permissions. Management and storage, and can include management of user-side user management and payment mechanisms. Multiple regional servers may be formed by hundreds of network nodes. Multimedia data can be stored in multiple blocks on different nodes, and can be mutually redundant and distributed.

After the local media segments capture and process the local multimedia clips, the integration is stored in a multimedia server (not shown), which can usually be a relational database, each of which has a time stamp. Therefore, it is possible to quickly obtain related multimedia clips through the mark.

In one of the embodiments, the regional server will use the multimedia content in the region, for example, using the side recording method to extract the TV channel and the network media content of the region, and then classify the multimedia content according to the characteristics of the multimedia content. With the mark, and the regional servers in different regions can also exchange records of classifications and marks.

In the media distribution system, the user terminal device 501 can log in to the system through the user login interface provided by the main server 503, and through the interface and the menu, the desired multimedia content can be selected, for example, a specific sport, a specific country. The media content, the self-interest, and the request generated via the interface will load various feature values that meet the specific requirements, and the main server 503 or the regional server then uses the media automatic editing method and the feature database and the content capture module. The media automatic editing device of the content classification module, the content re-creation module and the content broadcast module, import feature values, and compare the multimedia segments that meet the feature values, including shadows, sounds, characters or a combination thereof.

In the media distribution system, the main server 503 is connected to regional servers installed in various regions and countries, and each regional server will continue to record local multimedia content, including the TV channels and networks in the area, and a plurality of The regional servers (51, 52, 53) are also interconnected via the Internet 50 and are mutually redundant.

The main server 503 will obtain the records of the classification and marking of the plurality of regional servers (51, 52, 53) through the network. When receiving the request from the user terminal (the user terminal device 501), the request can also correspond to the image features. And the sound feature and/or the character feature, and may include the specific time information, and then the user server 501 and the corresponding regional server are connected according to the request of the user device 501 to compare the regional server where the corresponding multimedia content is located. The user device 501 directly loads the reworked multimedia segments through the Internet 50.

In this way, the media broadcast system can obtain a large amount of audio and video content, and the user can obtain the required multimedia clips through the network, and the content is compared and filtered by the features.

Fig. 6 is next a flow chart showing a method of generating a program table applied to the above-described media distribution system. The media broadcast system automatically obtains the multimedia clips of the local area by the local area server, and classifies, marks and automatically edits them to form a program list, which can be selected by each user terminal. The generation of the program list can be described with reference to the flow chart of FIG. 6.

The first step is as follows: S601, the regional server captures the multimedia content of the region, and performs classification and labeling through the media automatic editing technology, and each server can exchange the classification mark (step S603), thereby constructing a large and content-rich Media broadcast system.

The regional server will transmit the classification mark to the main server through the network (step S605), and the main server will then import the common classification mark of the user end (step S607), thereby correspondingly capturing the captured multimedia content and the client end. At the request, a program list is generated (step S609), and the program list is provided for transmission to the client (step S611).

The above media broadcast system refers to the flow of the embodiment of the media broadcast method as described in FIG.

The first step is to establish a connection between a main server and a plurality of regional servers in different regions, and then, in step S701, each regional server captures multimedia content of the region, including television programs and network media contents. It even includes analogies, digital broadcasts, audio and video information disseminated on various networks, etc. Through the integration, a diverse media database can be established.

Further, in step S703, each regional server performs classification and marking of the captured multimedia content according to the set features, wherein the automatically segmented multimedia segments include preset conditions provided by the system (71), and also refer to local users. Common conditions generated by attributes or their common content (72).

In step S705, each regional server will perform automatic media content editing or marking. Referring to the descriptions of FIG. 3 and FIG. 4 above, the multimedia clips matching various features may be clipped or marked, or the corresponding multimedia segments may be marked. Records form a database.

When the primary server receives the request signal from the client (step S707), the primary server obtains the features and requirements recorded in the request, and then polls the regional servers (step S709), and compares the regional server where the desired content is located. And obtaining, by each regional server, a segment or mark of the multimedia content that meets the request signal (step S711), and the main server re-arranges the video text content to form a new multimedia content (step S713), and broadcasts the content to issue a request. User terminal (step S715).

FIG. 8 shows the media broadcast process between the client, the main server and multiple regional servers mentioned in the disclosure.

Regional servers set up in different regions will capture the multimedia content of the region at any time, and classify and mark the features, and the regional servers (such as the first server and the second server shown) These classification and tag information can be exchanged with each other (steps 801, 802), and the first server and the second server respectively transmit the information to the host server (steps 803, 804).

After integrating the information of the multimedia content of each region via the main server, a program list provided to the client or a menu provided by the network interface may be established as described in FIG. 6.

The client can generate a request (step 805) to each type of multimedia player, communication device or computer system, and send it to the host server. Then, the main server sends a request message to the regional server than the feature data contained in the request, and sends a request message to the first server and the second server respectively according to steps 806 and 807. The first server and the second server respectively respond to the message (steps 808, 809).

The messages replied by the first server and the second server include whether there is content that meets the request of the client, and the host server responds to the client again (step 810), and the host server mediates the specific region that meets the request. The server, the client then confirms that the local server (in this case the first server) can transfer the content (step 811), and the re-made multimedia segment is transmitted by the regional server (step 812).

In summary, the automatic editing method and media broadcast technology proposed by the present disclosure, in particular, the process of automatically summarizing, re-manufacturing and reusing the media content, allows the user to capture, edit and edit in a manner that reduces labor costs. Organize a large amount of existing media content, and effectively provide the multimedia content that the user can obtain.

However, the above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Therefore, equivalent structural changes that are made by using the specification and the contents of the present invention are equally included in the present invention. Within the scope, it is combined with Chen Ming.

101. . . Original multimedia clip

103. . . Characteristic condition

105. . . Clip or mark a clip

107. . . New multimedia clip

20. . . Media automatic editing device

201. . . Content capture module

203. . . Content classification module

205. . . Feature database

207. . . Memory module

209. . . Content remanufacturing module

211. . . Input module

213. . . Content delivery module

31,41. . . Shadow

32,42. . . sound

33,43. . . Text

50. . . Internet

501. . . User terminal device

503. . . Master server

511. . . Area one

512. . . District II

513. . . Area three

51. . . First server

52. . . Second server

53. . . Third server

71. . . Preset condition

72. . . Common conditions

Steps S301 to S319 One of the media automatic editing processes

Step S401~S417 Media automatic editing process 2

Step S601 to S611 program list generation process

Steps S701 to S715 One of the media broadcast processes

Steps 801-812 Second of the media broadcast process

1 is a schematic view showing an embodiment of automatic media editing of the present invention;

2 is a block diagram showing an embodiment of a media automatic editing device of the present invention;

3 is a flow chart showing an embodiment of a method for automatically editing a medium according to the present invention;

4 is a flow chart showing the second embodiment of the automatic media editing method of the present invention;

FIG. 5 is a schematic diagram of an embodiment of a media broadcast system according to the present invention; FIG.

6 is a flow chart showing a method for generating a program table applied to the media broadcast system of the present invention;

FIG. 7 is a flowchart of an embodiment of a media broadcast method according to the present invention;

FIG. 8 is a second flowchart of an embodiment of a media broadcast method according to the present invention.

20. . . Media automatic editing device

201. . . Content capture module

203. . . Content classification module

205. . . Feature database

207. . . Memory module

209. . . Content remanufacturing module

211. . . Input module

213. . . Content delivery module

Claims (38)

A media automatic editing device includes: a feature database for recording one or more feature values; a content capture module for receiving one or more multimedia segments; and a content classification module for obtaining the received multimedia segments, Importing the feature value of the feature database and the attribute of the user end and the feature information collected by using, performing feature comparison, classifying the received multimedia segment; and a memory module for receiving or classifying the temporary storage The one or more multimedia segments; an input module for obtaining a request from one of the users; and a content re-manufacturing module, corresponding to requesting to obtain one or more multimedia segments, and reconstituting to form a new multimedia a segment; and a content delivery module that broadcasts a new multimedia segment corresponding to the request to the client that issued the request. The media automatic editing device according to claim 1, wherein the feature values recorded in the feature database include image features, sound features, and text features. The media automatic editing device of claim 2, wherein the feature value comprises a time information. The media automatic editing device of claim 2, wherein the image feature is an image color distribution of each frame in the multimedia segment. The media automatic editing device of claim 2, wherein the sound feature is a sound waveform, amplitude and frequency change. The media automatic editing device according to claim 2, wherein the The text feature is the subtitle of the multimedia clip. The media automatic editing device of claim 4, wherein the request corresponds to the image feature, the sound feature and the text feature, or a combination thereof. The media automatic editing device of claim 1, wherein the source of the multimedia segment comprises an image, a video, a sound file, a television channel, a network stream, and a camera. A method for automatically editing a medium, comprising: capturing one or more multimedia segments; importing one or more feature values and attribute information of the user end and feature information collected by using; performing media feature analysis, analyzing one of the captures or Characterizing a plurality of multimedia segments; comparing, according to a request, one or more multimedia segments conforming to the features recited in the request; temporarily storing the one or more multimedia segments conforming to the request to a memory; and Requesting one or more multimedia segments to form a new multimedia segment. The media automatic editing method according to claim 9, wherein the feature value is an image feature, a sound feature, a character feature, or a combination thereof. The media automatic editing method of claim 10, wherein the feature value comprises a time information. The media automatic editing method according to claim 10, wherein the image feature is an image color distribution of each frame in the multimedia segment. The media automatic editing method of claim 10, wherein the sound feature is a sound waveform, an amplitude, and a frequency change. The media automatic editing method according to claim 10, wherein the character feature is a subtitle of the multimedia segment. The media automatic editing method of claim 12, wherein the request corresponds to the image feature, the sound feature and the text feature, or a combination thereof. The media automatic editing method according to claim 9, wherein the source of the multimedia segment comprises an image, a video, a sound file, a television channel, a network stream, and a camera. The media automatic editing method according to claim 9, wherein the request is a client demand received through a network. The media automatic editing method according to claim 9, wherein after comparing one or more multimedia segments that meet the features described in the request, the one of the requests is captured in a clipping manner. Or a plurality of multimedia segments, combined and re-created into the new multimedia segment. The media automatic editing method according to claim 9, wherein after comparing one or more multimedia segments that meet the features described in the request, the marking is marked by one of the requests or a plurality of multimedia segments; the new multimedia segments are formed after the combined multimedia segments are combined. A media broadcast system includes: a main server, providing user management functions, a user login interface, a program menu, collecting popular keyword functions, and receiving a request from a client; A plurality of regional servers are connected to the main server through the Internet, and the regional servers capture the multimedia content in the region, and classify and mark according to the characteristics of the multimedia content, and exchange classification and markup between servers in each region. Recording, wherein the main server obtains the records of the classification and marking of the plurality of regional servers, and compares the corresponding multimedia content according to the classification of the plurality of regional servers and the marked records according to the request of the user end The regional server connects the client to the corresponding regional server. For example, the media distribution system described in claim 20, wherein the regional server is connected to the television channel of the region where the server is located and the regional network. For example, the media distribution system described in claim 20, wherein the plurality of regional servers are connected to each other through the Internet and are mutually redundant. The media broadcast system of claim 20, wherein the regional server comprises a media automatic editing device, the media automatic editing device comprises: a feature database, which records one or more feature values; Taking a module, receiving the multimedia content; a content classification module, obtaining the received multimedia content, importing the feature value of the feature database, performing feature comparison, classifying the received multimedia content; and a memory module For temporarily storing the multimedia content; a content re-manufacturing module, corresponding to requesting to obtain one or more segments in the corresponding multimedia content, and reconstituting to form a new multimedia content; A content delivery module broadcasts the new multimedia content that should be requested to the client. The media broadcast system of claim 23, wherein the feature values recorded in the feature database include image features, sound features, and text features. The media distribution system of claim 24, wherein the feature value comprises a time information. The media distribution system of claim 24, wherein the image feature is an image color distribution of each frame in the multimedia segment. The media distribution system of claim 24, wherein the sound feature is a sound waveform, an amplitude, and a frequency change. The media distribution system of claim 24, wherein the character feature is a subtitle of the multimedia segment. The media distribution system of claim 26, claim 27 or claim 28, wherein the feature corresponding to the request is one of the image feature, the sound feature and the text feature, or a combination thereof. A media broadcast method includes: establishing a connection between a main server and a plurality of regional servers setting different regions; each regional server extracts multimedia content of a region in which it is located; and classifying servers according to the set characteristics And the multimedia content captured by the mark; the records of the classification and the mark are exchanged between the servers of each area; the servers of each area transmit the records of the classification and the mark to the main server; The main server is configured according to the received classification and markup records and the classification marks commonly used by the individual users, thereby correspondingly selecting the captured multimedia content and the requirements of the individual client to generate a program list suitable for the individual client. And transmitting the program list to the individual client; the primary server receives a request generated by the individual client according to the program table; the primary server polls each regional server; the primary server obtains the request a segment or mark of the multimedia content; the main server re-forms a new multimedia content; and broadcasts the multimedia content to the client. For example, in the media broadcasting method described in claim 30, in which the television channels of the regions in which the servers are connected are connected to the regional network, a plurality of regional servers are connected to each other through the Internet, and are mutually redundant. The media broadcast method of claim 30, wherein the feature comprises an image feature, a sound feature, and a text feature. The media broadcast method of claim 32, wherein the feature comprises a time information. The media broadcast method of claim 32, wherein the image feature is an image color distribution of each frame in the multimedia segment. The media broadcast method of claim 32, wherein the sound feature is a sound waveform, an amplitude, and a frequency change. The media broadcast method of claim 32, wherein the character feature is a subtitle of the multimedia clip. The media broadcast method according to claim 34, claim 35 or claim 36, wherein the feature corresponding to the request is one of the image feature, the sound feature and the text feature or a combination thereof. The media broadcast method according to claim 30, wherein the regional server captures the content of the television channel in the area where the multimedia content of the region is located, and exchanges the network packet with the network of the region. .