JP2009225116A - Video recording device with network transmission function - Google Patents

Abstract

When transferring recorded moving image data, it takes time to transfer data other than the data related to the time range that the user thinks is really necessary, and the target video cannot be transferred promptly.
A keyword input unit 104, a keyword / weight correspondence dictionary 103, and a weight information recording unit 105 are provided, and a keyword input from the keyword input unit 104 during recording of a moving image file 100 is stored in the keyword / weight correspondence dictionary 103. A search is performed, and the corresponding weight is recorded in association with the time on the moving image file 100. When the moving image file 100 is transferred over the network, the moving image file 100 is divided into independent moving image files at regular intervals, and the moving image file 100 is transferred starting from the one with the largest total weight associated with each time. When all transfer moving image files have been transferred, a playlist in which the transferred files are rearranged in order of recording time is generated by the playlist generation means 109 and transferred.
[Selection] Figure 1

Description

  The present invention relates to a video recording apparatus such as a video camera, and more particularly to a video recording apparatus that transfers a recorded video to a remote device by network transmission.

  In recent years, as represented by the spread of WiMAX, wireless LAN hotspots in towns are increasing. This makes it realistic to upload digital data such as photographs taken with a digital camera from a wireless LAN hotspot to a home personal computer, server, online service, or the like (see, for example, Patent Document 1).

In the conventional example, a wireless transmission device such as a mobile phone and a position detection device such as GPS are connected to the digital camera, and a position where transmission is possible with the wireless transmission device is detected from the position detection device, and a position where transmission is possible has been reached. At this time, the wireless transmission device is controlled to transmit the imaging data recorded on the memory card to a desired destination. That is, when the digital camera presented in the conventional example is brought into the wireless LAN hotspot area, the captured data can be automatically uploaded to the Internet.
Japanese Patent Laid-Open No. 2000-156813

  In the conventional example, the data handled is still image data. However, when moving image data is handled in the transmission system presented in the conventional example, the following problems occur in operation.

  In general, moving image data has an extremely large data size compared to still image data. Therefore, when a file storing data is wirelessly transferred, a very long time, which is substantially proportional to the data size, is required.

  On the other hand, the moving image data can be regarded as a series of frame images associated with the shooting time, but the moving image data file format is either MPEG-2 (ISO / IEC 13818) or DV (IEC 61834). In a typical video file, the later data tends to be stored later in the file.

  Due to the above features, if the data of the part of the video file to be transferred that the user wants to transfer is located behind the video file, a long time from the start of the transfer of the video file is less important There is a problem that a side that is spent on data transfer and receives a moving image file through a network cannot easily access a necessary part, that is, it takes time until a part important for a user can access a part.

  For example, this problem becomes a very serious problem when a video file of a news video recorded by a camera recorder at a broadcasting station is to be transferred to the main station as soon as possible using a wireless network.

  The present invention has been made in view of such a problem, and an object of the present invention is to provide a video recording apparatus that preferentially transfers a video portion important for a user and allows quick access on the receiving side.

  In order to solve the above-described problems, a video recording apparatus according to the present invention includes a video input unit that inputs video, an input unit that receives input from a user, and a weight dictionary that holds weight information corresponding to input contents received by the input unit. Weight collating means for determining weight information from the input content with reference to the weight dictionary, recording means for recording the weight information in association with a designated position in the video, and division for dividing the video into a plurality of partial videos Priority order between partial videos based on weight information associated with each of the partial videos, and reproduction order information creating means for creating the reproduction order information indicating the reproduction order of the video before division for the divided partial videos And sending means for sending the partial video in the descending order of priority and sending the playback order information.

  The video recording apparatus of the present invention has the above-described configuration, and even when transferring a moving image file, the transfer starts from the video part important for the user, and at the stage when all the transfer is completed, the entire video of the transfer source Since the image can be reconstructed at the transfer destination, it is possible to realize efficient transfer of the moving image file for the user.

  FIG. 1 is a block diagram showing a configuration of a video recording apparatus according to an embodiment of the present invention.

  In the video recording apparatus of the present invention, the uncompressed moving image data output from the imaging unit 113 is compressed into a moving image format such as MPEG-2 or DV by the moving image compressing unit 110 and is stored as a moving image file 100 on the recording medium 116 through the file system interpretation unit 114. To be recorded. The file system interpretation unit 114 manages the arrangement of files on the recording medium 116 using, for example, a FAT file system (ISO / IEC 9293).

  At the same time, the video recording apparatus of the present invention comprises a keyword input means 104, a keyword / weight correspondence dictionary 103, a keyword / weight collation means 112, and a weight information recording means 105 in order to record the weight correspondence recording file 101 on the recording medium 116. . In addition, a moving image transmitting unit 102 is provided for transferring the moving image file 100 through the wireless transfer unit 108, and the moving image transmitting unit 102 includes a file dividing unit 107, a transmission order determining unit 106, and a playlist generating unit 109.

  The video recording apparatus of the present invention is a camera recorder using a semiconductor memory as a recording medium, for example. In the following description, a camera recorder using a semiconductor memory as a recording medium will be described as an example.

  First, a method for creating the weight correspondence recording file 101 will be described in detail.

  The keyword / weight correspondence dictionary 103 is a table in which one-to-one correspondence between words representing persons, objects, and events that can be reflected in the content to be photographed, and numerical values representing the degree of importance determined in advance for each word. This table is created and registered by the user before shooting.

  FIG. 4 is a diagram showing an example of the keyword / weight correspondence dictionary 103. In this example, importance “3” is associated with the word “PRESIDENT”, and importance “2” is associated with the word “GUEST”.

  The keyword input unit 104 is, for example, a microphone that picks up a user's voice, a touch panel or a keyboard that receives character input or selection input from the user.

  The user inputs a keyword about a subject currently being photographed through the keyword input unit 104 in real time while recording moving image data. For example, if the appearance of “PRESIDENT” is within the shooting range, “PRESIDENT” is input, and if “GUEST” appears, “GUEST” is input.

  The keyword / weight collation unit 112 collates the input content from the keyword input unit 104 with the keyword / weight correspondence dictionary 103 and notifies the weight information recording unit 105 of the corresponding importance. Upon receiving the notification, the weight information recording means 105 shows the correspondence between the received weight information and the time information at that time on the same recording medium 116 on which the moving picture file 100 is recorded, through the file system interpretation means 114. 100 is written as a weight correspondence recording file 101 associated with 100. The weight information recording unit 105 generates the weight correspondence recording file 101 at the start of recording of the moving image file 100, and writes the correspondence between the weight information and the time information at that time each time the user inputs a keyword.

  FIG. 7 is a diagram showing a method of associating the moving image file 100 and the weight correspondence recording file 101. In the figure, the same reference numerals are used to indicate the same components as those in FIG.

  In the present embodiment, the name of the weight correspondence recording file 101 is determined based on the name of the moving image file 100, thereby associating the two. For example, a moving image file during a recording operation is stored in the recording medium 116 with CONTENTS / VIDEO / 00017J. In the case of a file named MXF, the weight information recording means 105 adds the character “_w” to the same name as the moving image file 100, and the extension is in accordance with the description method of the meta information, that is, “.xml”. The weight correspondence recording file 101 changed to "" is created in a folder dedicated to meta information. In this example, CONTENTS / VIDEO / 00017J. As a weight-corresponding recording file 101 corresponding to the moving image file 100 named mxf, META / 00017J_w. A file called xml is created. By adopting the above naming method, it is easy to see that the moving image file 100 and the weight correspondence recording file are associated with each other because the main part of the file name is the same. As the association between the moving image file 100 and the weight-corresponding recording file 101, an ID common to each file, for example, an ID assigned when recording the moving image file 100 is included in the weight-corresponding recording file 101. Also good.

  FIG. 6 is a diagram showing an example of description contents of the weight correspondence recording file 101 in the video recording apparatus of the present invention.

In this example, XML (extensible Markup Language [JIS X 4159: 2002]) is used for the description of the weight correspondence recording file 101. Here, the <Weights> tag indicates that the correspondence between detected weights and time is listed below, and the <Weight> tag indicates the contents thereof. The contents of the first <Weight> tag in the example shown in FIG.
<Weight time = "00:11:21" weight = "1" word = "AUDIENCE"/>
Is the keyword “AUDIENCE” input from the user at the time 00:11:21 in the moving image file 100 associated with the weight correspondence recording file 101, and the keyword recorded in the keyword / weight correspondence dictionary 103. Represents that the weight of “1” was “1”. Hereinafter, <Weight> tags are enumerated by the number of keywords detected by the keyword / weight collating unit 112.

  Through the above operation, the weight correspondence recording file 101 associated with the moving image file 100 and simultaneously associated with the time in the moving image file 100 can be created on the recording medium 116 at the same time when the moving image file 100 is recorded. .

  Next, a method for transferring a moving image to a network in a video recording apparatus provided with the network transmission system of the present invention will be described in detail with reference to FIGS.

FIG. 2 is a diagram illustrating an example of the relationship between the moving image file 100 and the weight correspondence recording file 101. The weight correspondence recording file 101 associates W _t0 , W _t1 , W _t2 ,... With each of the times t ₀ , t ₁ , t _{2 ,.}

  Prior to sending the moving image file 100 to the network, the transmission order determination unit 106 first reads the weight correspondence recording file 101 from the recording medium 116 via the file system interpretation unit 114.

  FIG. 3 is a diagram showing a transmission order determination operation performed by the transmission order determination unit 106 based on the contents read from the weight correspondence recording file 101.

  The transmission order determination unit 106 divides the moving image file 100 into time ranges for each time Δt, and calculates the sum of the associated weights for each time range. Note that Δt is an arbitrary time determined in advance by the user or the system. When a method using GOP (Group of Pictures) such as MPEG-2 is used for compressing video data, the time corresponding to 1 GOP is used. Integer multiple. Further, the transmission order determination unit 106 rearranges the time ranges in the order of the largest weights, and creates the transmission order list 300.

  The operation of the transmission order determination means 106 will be specifically described taking the case where the weight correspondence recording file 101 shown in FIG. 6 is given.

  FIG. 10 is a diagram showing the relationship among the moving image file 100, the weight correspondence recording file 101, and the transmission order list 300. FIG. 10A shows the relationship between the moving image file 100 and the weight correspondence recording file 101. If the user sets the division time unit Δt as 10 minutes (00:10:00), the sum of the weights associated with each time range is the value shown in FIG. For example, in the time range 1001 from 0:50:00 to 1:00:00, there are two associated weights, both of which are 3, so the total weight in this time range 1001 is 6. Similarly, the total weight is obtained for other time ranges.

  Next, the transmission order determination means 106 sorts in descending order of the total weight calculated by the above procedure, and creates a transmission order list 300 (FIG. 10C). In the case of the example of FIG. 10, the time range 1001 is the top of the transmission order list 300 because the total weight of the time range 1001 is as large as 6.

  When the total weights associated with the time ranges are the same (both 0 in this case) as in the time range 1002 and the time range 1003, the transmission order list 300 has the largest average weight including the adjacent time range. Here are the top ranks. If it is not possible to determine only by the average value including the adjacent neighbors, the adjacent time range included in the average calculation is increased until it can be determined. For example, in the time range 1002 and the time range 1003, the average value of the weights associated with the three time ranges, that is, the time range immediately before and the time range immediately before the time range, is the same as 0.33. Calculate the average value of the weights associated with the five time ranges added to the time range two before and after the respective time range, the time range 1002 is 0.20, the time range 1003 is .0. 40, and it can be determined that the time range 1003 is listed higher in the transmission order list 300 than the time range 1002.

  When the transmission order list 300 has been created, the transmission order determination unit 106 notifies the file division unit 107 of the time ranges listed in the transmission order list 300 in order from the top. In the example of FIG. 10, first, the time range 1001 from 0:50:00 to 10:00, which is the highest priority section, is notified, then the range from 1:10:00 to 1:20:00, and the next Is informed of the range from 0:40:00 to 0:55:00.

  Next, the operation of the file dividing unit 107 notified from the transmission order determining unit 106 from a certain time range nΔt to (n + 1) Δt (n is an arbitrary integer) will be described.

  The file system dividing unit 107 accesses the moving image file 100 on the recording medium 116 via the file system interpreting unit 114 and cuts out and reads the video data of (n + 1) Δt from the notified time range nΔt. Further, the file system dividing unit 107 performs the following two processes on the video data read in the above procedure.

  First, as a first process, if necessary, a header and a footer conforming to the video compression method used in the moving image file 100 are added to the cut out video data.

  For example, when the moving image file 100 uses a complete algorithm for each frame such as the DV system (IEC 61834), this procedure is not necessary. On the other hand, when the video data is composed of MPEG-2 TS (Transport Stream: ISO / IEC13818-1: 2000), this processing is necessary. Hereinafter, an example of this case will be described.

  FIG. 8 is a diagram showing a configuration when the moving image file 100 is in the MPEG2-TS format and an example of a header that needs to be rewritten. Since MPEG-2 is used for compressing video data as described above, Δt is an integral multiple of the time corresponding to 1 GOP, and the start point nΔt and end point (n + 1) Δt of the time range are just the GOP boundary. Make sure they match. The file system dividing unit 107 extracts, as an ES (Elementary Stream) file 800, a video regarding the range from the time range nΔt to (n + 1) Δt notified from the transmission order determining unit 106 from the moving image file 100. This can be easily obtained by removing the TS header and the PES header from the TS stream including the corresponding time range and cutting out the ES in the required time range. In order to make the ES file 800 an independent TS file again, the file dividing unit 107 divides the ES file 800 at an arbitrary size, adds a PES header, and converts the file into a plurality of PES packets 801. At this time, it is necessary to write the size used for the division as a PES packet length (PES_Packet_Length) in the PES header. Further, the file dividing unit 107 further divides the PES packet 801 by a predetermined size, adds a TS header, and creates a plurality of TS packets 803. The data in which these TS packets are continued becomes a TS file. At this time, it is necessary to write a TS packet serial number (Continuity_Counter) in the TS header.

  Therefore, in order to divide the moving image file 100 within the time range notified from the transmission order determining unit 106, the file dividing unit 107 generates TS header and PES header values according to the extracted time range and adds them to the data. There is a need to.

  Next, as a second process, an MXF header and an MXF footer are added to the data (the video data read when the first process is unnecessary). MXF (Material eXchange Format: SMPTE 377M-2004) is a file exchange format and image compression format standardized by the Society of Motion Picture and Television Engineers (SMPTE). It is suitable for exchanging multimedia information as a file because it is a format that can store video, audio, and metadata in a single file.

  FIG. 5 is a diagram showing an example of the structure of the MXF file. In the file header, metadata (header metadata) defined by MXF and an index table are multiplexed. In the file body, the data acquired in the first process is stored as it is. In the file footer, the same index table as the file header is multiplexed.

  Since the header metadata area of the MXF file is defined so that it can be independently expanded on the user side, in this embodiment, the UMID (Unique Material Identifier: SMPTE 330M) and transmission order of the moving image file 100 before division are provided. A list of keywords used by the determining means 106 for adding the weights is added as meta information as metadata of the MXF file. For example, for video data divided corresponding to the time range 1001, the same UMID, keywords “PRESIDENT” and “SPEECH” as the original moving image file 100 are added as metadata.

  The file dividing unit 107 divides the moving image file 100 into (n + 1) Δt from the time range nΔt notified from the transmission order determining unit 106, and further wirelessly transfers the data obtained through the above two processes as a moving image file for transfer. Transfer to means 108. The wireless transfer means 108 uploads the received transfer video file on the network. In the present embodiment, the file name at this time is obtained by adding a serial number in the transfer order to the end of the file name of the moving image file 100. For example, in the example of FIG. Since the file name is mxf, the file name to be transferred is 00017J-1. mxf, 00017J-2. mxf,...

  By performing the above two processes on each of the cut-out video data, the transfer video file to be transferred becomes a video file that can be played back independently. The file can be viewed alone. Further, since the transfer order of each time range is determined based on the weight correspondence recording file 101, the receiving side can view the videos as soon as they are obtained in descending order of importance. In addition, if you use an application that supports the interpretation of metadata that has been uniquely expanded, you can acquire the keywords that are the basis for weighting by reading the metadata of the received video file for transfer, The UMID of the file 100 can be acquired.

  By repeating the operation of the file dividing unit 107 described above in order for all the time ranges notified by the transmission order determining unit 106, substantially the entire video included in the moving image file 100 is displayed in a highly important scene. Can be transferred from.

  Further, the transmission order determining unit 106 notifies the play list generating unit 109 of the transfer order of the entire time range of the moving image file 100 when the file dividing unit 107 has been notified of all the time ranges of the moving image file 100. Based on the notified contents, the playlist generating means 109 creates a playlist as playback order information for rearranging the divided videos in the order of recording time and playing them back.

  FIG. 9 is a diagram showing an example of a playlist file generated by the playlist generation unit 109 based on the transmission order list 300 shown in FIG.

  The transmission order determination unit 106 notifies the playlist generation unit 109 of the transmission order list 300 shown in FIG. The notified data always includes the correspondence between the start time for each time range divided using the file dividing means 107 and the order in which they are transferred.

  Based on this list data, the playlist generating means 107 rearranges the transfer moving image files in the order of the start times of the respective time ranges as reproduction order information, and creates a playlist file as shown in FIG. In this embodiment, SMIL 1.0 (Synchronized Multimedia Integration Language 1.0: W3C Recommendation) is adopted as a description format of the playlist. SMIL is a video processing language based on XML.

  The <SEQ> tag in SMIL indicates that the elements listed in the tag are sequentially and sequentially reproduced. That is, in the example of FIG. 9, first, 00017J-8. mxf is played, followed by 00017J-5. mxf then 00017J-7. This indicates that the data should be reproduced in the order of mxf. As can be seen from a comparison with FIG. 10C, this indicates that the time range notified from the play list generating means 109 is rearranged in the originally photographed order and reproduced.

  The playlist generation unit 109 transfers the playlist file created by this procedure to the wireless transfer unit 108. The wireless transfer means 108 sends the transferred playlist file over the network. On the receiving side of the playlist file via the network, the multiple transfer video files and playlist files that have been transferred in order of priority based on the total weight are handled as a set, and finally the recording of the source camera recorder is performed. It is possible to receive a video that is completely equivalent to the moving image file 100 recorded in the medium 116.

  As described above, the video recording apparatus of the present invention can quickly transfer the captured video from the highly important part to the network, and at the same time, restore the video that is completely equivalent to the transfer source at the transfer destination. become able to.

  In the present embodiment, it has been described that the keyword input using the keyword input unit 104 is performed at the same time as shooting, but this is because the moving image file 100 once recorded on the recording medium 116 is reproduced and the reproduced video is displayed. You may make it input, confirming.

  In the present embodiment, the keyword input unit 104 is used to input a keyword. However, the method for inputting a keyword from the keyword input unit 104 is not limited to this. That is, a voice input means such as a microphone picks up sound heard in the surroundings during shooting, and a keyword may be extracted by performing voice recognition on this voice input, or image recognition from the video being shot. Keywords may be extracted using technology.

  In any case, it is sufficient that the keyword input is made before the moving image file 100 is transferred.

  In the present embodiment, the keyword / weight correspondence dictionary 103 includes only words including natural language words as keywords, but the keywords may be user actions themselves. That is, the camera recorder includes several buttons as the keyword input unit 104, the keyword / weight correspondence dictionary 103 includes a table in which the pressed button is associated with the importance, and the keyword / weight collation unit 112 is a keyword The weight information recording unit 105 may be notified of the importance corresponding to the pressed button with reference to the weight correspondence dictionary 103.

  In the present embodiment, only the format based on XML as the weight-corresponding recording file 101 and the playlist is given, but any format may be used as long as it provides the same function.

  In the present embodiment, the moving image file 100 is divided into a moving image file for transfer by dividing the moving image file 100 within a predetermined time range. However, the moving image file 100 is not divided every predetermined time, but divided according to other rules. May be. For example, the file may be divided by the time when the keyword is input from the keyword input unit 104 (time when the weight information is associated). In this case, weight information is associated with the head of each time range, and one weight information is associated with each time range. By doing in this way, it is possible to transmit an important part frequently input with keywords with a finer priority. Moreover, you may use together with the method divided | segmented for every predetermined time as demonstrated in this Embodiment. For example, it is possible to divide every predetermined time and also divide the time when a keyword is input. Moreover, although it divides | segments by the time when keyword input was made, when there is no keyword input more than predetermined time, you may make it divide | segment by the time.

  In the transfer video file naming rules, only the example in which the transmission order is added to the end of the file name is given in the present embodiment. However, the information on the corresponding time range is added to the file name. There may be only information relating to the time range. For example, if the transfer moving image file to be transferred first is in the time range starting from 0:50:00, that is, 5Δt as shown in FIG. 10C, 00017J-5-1. You may name it like mxf. In this example, “5” corresponding to the eighth character of the file name represents a serial number in the order of time range, that is, reproduction order information that is the order in which the image was originally taken, and “1” corresponding to the tenth character has been described in the present embodiment. Represents the order of transmission. In this case, when the transfer video file is received via the network, it is possible to restore the transfer source video file 100 by sorting using the file names and reproducing in that order. However, since it is considered that only the functions of the file dividing means 107 and the playlist generating means 109 are integrated, there is essentially no difference from the present invention. .

  The video recording apparatus provided with the wireless transmission system according to the present invention is useful for the purpose of efficiently uploading recorded video data to the Internet, and is particularly suitable for a system for quickly transferring necessary video to a remote place. ing.

The block diagram which shows the structure of the video recording apparatus in embodiment The figure which shows the example of the relationship between a moving image file and a weight corresponding | compatible recording file The figure which shows the operation of transmission order decision The figure which shows the example of keyword / weight correspondence dictionary Diagram showing an example of the structure of an MXF file The figure which shows the example of the description contents of the weight correspondence record file The figure which shows the method of correlating the movie file and the weight correspondence record file The figure which shows the example of the structure of MPEG2-TS, and the header which needs rewriting The figure which shows the example of the playlist file A diagram showing the relationship between a movie file, weight-corresponding record file, and transmission order list

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Movie file 101 Weight correspondence recording file 102 Movie transmission means 103 Keyword / weight correspondence dictionary 104 Keyword input means 105 Weight information recording means 106 Transmission order determination means 107 File division means 108 Wireless transfer means 109 Playlist generation means 110 Movie compression means 112 Keyword / weight collation means 113 Imaging means 114 File system interpretation means 116 Recording medium 300 Transmission order list 1001 Maximum weight total time range 1002 Time range with zero total weight (1)
1003 Time range with zero total weight (2)

Claims (5)

Video input means for inputting video;
Input means for receiving input from the user;
A weight dictionary holding weight information corresponding to the input content received by the input means;
Weight collating means for determining the weight information from the input content with reference to the weight dictionary;
Recording means for recording the weight information in association with a designated position in the video;
Dividing means for dividing the video into a plurality of partial videos;
Reproduction order information creating means for creating reproduction order information indicating the reproduction order of the video before division for the divided partial video;
Sending means for deciding a priority between the partial videos based on the weight information associated with each of the partial videos, sending the partial videos in order from the highest priority, and sending the reproduction order information; ,
A video recording device comprising: The weight dictionary holds a keyword and weight information corresponding to the keyword,
The video recording apparatus according to claim 1, wherein the input unit receives a keyword registered in the weight dictionary as the input content. The video recording apparatus according to claim 1, wherein the division unit records information for specifying a video before division as additional information in the partial video to be generated. The video recording apparatus according to claim 1, wherein the dividing unit records the input content that is a basis for determining the priority order of the partial video as additional information in the partial video to be generated. 2. The recording unit according to claim 1, wherein the recording unit records the position of the video input from the video input unit and the weight information based on the input content in association with each other when the input unit receives the input content. Video recording device.