JP2005301641A - Video photographing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To recover data in case of an abnormal end such as removal of a medium during recording or sudden power outages, thereby rendering the data reproducible. <P>SOLUTION: A file format detection means 241 detects the file format of any failed file. Based on the format detected, an effective data length detection means 244 detects image and voice data in portions of the failed file already recorded. An essence file metadata recovery means 245 corrects the file to adapt to a file format in accordance with the effective data length detected. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention is a video that enables the material to be restored and played back again in the event of an abnormal termination such as suddenly removing the recording medium or turning off the power while recording a material such as video or audio on the recording medium. The present invention relates to a photographing apparatus.

  In recording devices such as video cameras and digital still cameras, and recording devices such as disk recorders, in order to record data on a randomly accessible recording medium such as a memory card, hard disk, or optical disk, the data is filed from the viewpoint of usability. It is common to record as In order to record a file on a recording medium, a module called a file system for managing files on the recording medium is used. Since the operation of the file system is complicated, it is common to mount an operating system (OS) on the photographing apparatus or recording apparatus and to implement the file system as software.

  There is FAT (Non-Patent Document 1) as a file system adopted in a memory card. The outline will be described with reference to FIG. A physical drive 600 such as a memory card or a hard disk is usually composed of a master boot record 102 and one or more logical drives 101 placed at the head of the area. One logical drive 101 is formatted with one type of file system. In the case of the FAT format, a system area such as a partition boot sector 103, a file allocation table (FAT) 104, a FAT backup 104b, and a root directory entry 105 is arranged from the top of the area, and a user data area 106 is arranged behind the system area. The In the partition boot sector 103, information necessary for starting the partition such as the number of sectors of the partition is recorded. File recording information is arranged in the FAT 104. In the user data area 106, file data itself is placed. Files are recorded in units called clusters (usually 4 KB to 32 KB).

  In the root directory entry 105, information on a file or directory placed in the root directory is recorded as an entry. In the hierarchical directory, all files and directories are recorded as directory entries 107 as shown in FIG. FIG. 2 shows FLOWER02. A directory entry of a file having a file name of AVI is shown. Thus, in the FAT format, a file is recorded in an 8.3 format with a file name of 8 characters and an extension of 3 characters. The directory entry 107 includes the first cluster number 108 in which file data is placed. The file data can be accessed by the head cluster number 108.

  The file system adopted for the DVD is UDF (Universal Disk Format) defined by the US OSTA (Optical Storage Technology Association). In UDF, instead of directly managing directory and file management information as actual recording positions (real addresses), a file identification descriptor (FID; File Identifier Descriptor), a virtual allocation table (VAT; Virtual Allocation Table), and a file entry It is managed indirectly using an ICB (FileEntry Information Control Block) or the like.

VAT is a technique for handling as if data can be updated or reproduced by accessing an arbitrary address on the medium when recording on the recording medium by the additional recording method. A real address actually recorded and a virtual address corresponding to the real address are managed as a correspondence table.

  In UDF, a group of logical sector strings in which data is written / read is referred to as an extent, and the start address of each extent is held as file information and additionally written together with file system information to be updated.

When a file is recorded on a removable recording medium using a photographing device, the file remains in an illegal state due to sudden removal or battery stop, etc., so that unreproducible data remains on the recording medium. Become. Many methods for repairing files in a file system in the event of such a failure have been proposed. Patent Document 1 proposes a method of recording a code of an erased file in a FAT directory entry and erasing fragment data remaining at the time of failure and an unerased area remaining at the time of file erasure. Patent Document 2 shows file restoration for UDF. When data recording is interrupted, the extent file name, file directory information file entry ICB, VAT, etc. are recorded in a predetermined area as restoration file information, A method of repairing a file based on the repair file information is proposed. Patent Document 3 proposes a method of recording program recording history information in a system area different from the user area, and performing restoration in order from the top address of a program file that has not been recorded.
JP-A-9-252447 JP 2003-169293 A JP 2001-143439 A Standard ECMA-107: Volume and File Structure of Disk Cartridges for Information Interchange

  However, since such a conventional video photographing device focused only on the restoration of the file system, it is possible to repair the consistency of the file format (make it possible to read and write as a file), but the video contained in the file Data and audio data formats could not be restored (so that they can be played back as video data or audio data). In addition, when recording audio and video data in separate files, it has not been possible to restore them with consistency.

  In view of the above problems, the video photographing apparatus of the present invention restores data even in the case of abnormal termination such as sudden removal of the recording medium or power failure during recording of video and audio material data on the recording medium. The purpose is to enable playback again.

  In order to solve the above-described problem, the video shooting apparatus of the present invention includes a file format detection unit that detects a file format of an abnormal file whose recording is interrupted, and an abnormal file based on the format detected by the file format detection unit. Effective data length detecting means for detecting the length of data that has been recorded among the recorded data, and metadata repair for correcting the metadata of the abnormal file based on the effective data length detected by the effective data length detecting means Means.

  According to the video photographing apparatus of the present invention, even when an abnormal end occurs, the data length of the recorded portion of the material can be detected and repaired, and the material can be reproduced.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
(Embodiment 1)
In the embodiment described below, the imaging apparatus records video, audio data, and the like as separate files in consideration of compatibility with editing. That is, a plurality of files are generated simultaneously at the time of shooting. The video and audio files generated at the same time are collectively called a clip. In this clip, in addition to video and audio, metadata attached to the clip is recorded as a separate file. This is called clip metadata. In addition to this, a metadata file such as a thumbnail image may be associated with the clip. A video or audio file that is a component of a clip is called an essence file.

  Metadata is data that describes information about data. The metadata related to a clip includes data necessary for reproducing a clip such as a codec type and a sampling rate, and data necessary for searching for a clip such as a photographer and a photographing time. The clip metadata of the present embodiment is recorded as an XML format file as shown in FIG. The EssenceList 201 is metadata necessary for reproduction, and the ClipMetadata 202 is metadata used for searching and the like.

  The essence file is recorded as an MXF file (see SMPTE377M). The MXF file has a format as shown in FIG. 4 and is roughly divided into a header part 210, a body part 211, and a footer part 212. The header unit 210 and the footer unit 211 record metadata such as a codec type, and the body unit 211 includes audio and video data. The video data is recorded in an intra-frame compression format called DIF data (compressed data defined by SMPTE 314). In the DIF, the same audio data as that recorded in the audio file can be multiplexed and recorded in the frame. That is, in the video data file 213, the same data as the audio file 214 and the audio file 215 is multiplexed and recorded on the video as audio data A216 and audio data B217.

  FIG. 5 shows a flow in which the image capturing apparatus of the present embodiment captures an image. First, the file storage location management means 220 determines the file name of the file to record the video, audio and metadata based on the directory structure shown in FIG. 6, and the clip metadata creation means 221, video file creation means 222, audio The file creation unit 223 is notified of the file name to be recorded. The file names in FIG. 6 are Clip1_0. mxf and Clip1. A clip name (Clip 1 in FIG. 6) is included in a file such as xml.

The clip metadata creation means 221 is a means for creating a clip metadata file. Information such as a codec type and a sampling rate that is determined before shooting is recorded in the metadata file at the time of shooting, and the clip length or The shooting end time or the like is not recorded and is invalid. For example, as shown in FIG.
<Duration> 0 </ Duration>
As a result, a clip length of 0 is recorded. Since a clip length of 0 cannot be specified in the XML schema, this can be used to determine that the clip is abnormal. The video file creation unit 222 and the audio file creation unit 223 record information determined before shooting in the header portion 210 of the essence file.

After making these preparations, video and audio are captured and data is recorded on a recording medium. The video captured by the video input unit 224 is compressed by the video file creation unit 222 and recorded on the recording medium 227 through the recording unit 226. Similarly, the voice captured by the voice input unit 225 is recorded on the recording medium 227 through the voice file creation unit 223 and the recording unit 226.
When the shooting is stopped, the video input unit 224 and the audio input unit 225 interrupt the capturing of the video and audio, respectively, and the clip metadata generation unit 221, the video file generation unit 222, and the audio file generation unit 223 perform the shooting end process of the clip. Do. The video file creation unit 222 and the audio file creation unit 223 create the footer unit 212 of the MXF file, write metadata such as the clip length and the shooting end time, and are not recorded in the header unit 210 before shooting. Write data and close the file. The clip metadata creation unit 221 writes similar information to the metadata file.

  If the recording medium 227 is unplugged before the processing at the time of stopping the shooting is completed, or if the device stops due to abnormal termination or the like, the abnormal video data file, audio data file, clip length, An incomplete clip with no shooting end time is created.

  A file constituting an abnormal clip may not be able to read data, that is, a file system error may occur. In order to repair such an abnormality in the file system, the method disclosed as the prior art is used, so that the file can be read and written normally and the data up to the time of shooting can be referred to.

  However, even if such a file system is repaired, the clip information remains inaccurate and can be read and written as a file, but still cannot be reproduced as video or audio on a playback device or the like. There is also a problem that reading fails because the metadata does not match the XML schema.

  A method for repairing such a defect will be described with reference to FIG. First, the abnormal clip detection means 240 refers to the effective data length (corresponding to Duration of FIG. 7) in the clip metadata, and recognizes that a zero is an abnormal clip as shown in FIG. The file format detection means 241 compares the format of the essence file recorded in the metadata with the format written in the header of the essence file, and if they are the same, uses it as the file format of the essence file. If they are different, the format confirmation means 242 is instructed to confirm the format to the user. The format confirmation means 242 allows the user to specify a format from the format list 280 as shown in FIG. The video playback means 243 performs playback in the designated format on the screen 281.

  The effective data length detection unit 244 sets the effective data length of each file based on the detected or designated format. The effective data length is the length of data that has been correctly recorded among the data included in the abnormal file. For example, in order to improve the performance of the writing process, when an abnormal termination occurs in a photographing device that writes essence data in units of 32 KB, the essence file has a size that is an integral multiple of 32 KB. The video data file length is 18,874,368 bytes (= 18 MB = 32 KB × 576), the audio data file length is 524,288 bytes (= 512 KB = 32 KB × 16), and the header size is 32,768. The effective data length detection method of the effective data length detection means 244 will be described taking the case of bytes (32 KB) as an example.

The effective data size of the video data is 18,841,600 bytes by subtracting the header size from all data according to the following formula.

18,874,368 (bytes)-32,768 (bytes)
= 18,841,600 (bytes)

If the format detection means 241 detects that this data is recorded in DVCPRO (see SMPTE 306M), the data size of one frame of DVCPRO is 120,000 bytes. Is normally recorded, and it can be seen that the recording ended abnormally during the recording of the 158th frame.

18,841,600 (bytes) ÷ 120,000 (bytes / frame)
= 157.001333 (frame)

On the other hand, the effective size of the audio data is 491,520 bytes by subtracting the header size from all data according to the following formula.

524,288 (bytes)-32,768 (bytes)
= 491, 520 (bytes)

If it is detected by the format detection means 241 that this data is 16 bits / sample (= 2 bytes / sample) of 48K sampling, data corresponding to 5.12 seconds is stored according to the following equation. I understand that.

491,520 (bytes) ÷ 2 (bytes / sample)
÷ 48,000 (samples / second) = 5.12 (seconds)

If this is converted into a frame at a frame rate of 29.97 frames / second, it can be seen that 153 frames of data are recorded and the recording ends abnormally during the recording of the 154th frame.

5.12 x 29.97 = 153.4464 (frame)

As described above, it can be detected that the effective data length of the video data is 157 frames and the effective data length of the audio data is 153 frames.

  The essence file metadata restoration unit 245 sets the last 1,600 bytes and 1,430 bytes of the video data and audio data as invalid data and writes a footer after the invalid data. Then, the essence file is restored by recording the number of frames obtained above in the header and footer of each file of video and audio.

  The effective data length adjusting unit 246 extracts the shortest of the essences and sets it as the effective data length of all material files associated with the clip. When the video file of the previous example is 157 frames and the video file is 153 frames, 153 frames are adopted as the number of frames of the clip and recorded in Duration in the clip metadata file.

  A UMID (see Universal Material Identifier, SMPTE 330M) is written on the DIF in a multiplexed frame. The UMID is an identifier that can uniquely identify a material before editing globally, and includes a basic UMID 250 that is a globally unique identifier and a source pack 251 that represents metadata (FIG. 10). The basic UMID 250 includes a universal label 252 corresponding to a key in the KLV (Key-Length-Value SMPTE 336M) format, a valid data length 253, an instance No. 254, and a material No. 255 created from a time, a random number, a device ID, and the like. Source pack 251 can contain metadata for time 256, location 257, country 258, organization 259, and user 260. The final frame metadata extraction unit 247 extracts the UMID time information from the DIF corresponding to the final frame of the clip when the UMID source pack is included in the DIF, and sets it as the shooting end time. Record in metadata file.

  As described above, the clip metadata and the essence file are in a state equivalent to normal completion and can be reproduced. The repaired clip can be used for editing and searching.

  Although the file format of the present embodiment has been shown as comprising a header, body, and footer, either or both of the header and footer may be omitted. Although the DIF format is taken as an example of the compression method, an inter-frame compression method such as MPEG may be used. In the case of MPEG, the effective data length is not uniquely obtained from the size, but is obtained by a method such as examining the first identifier (00001B8) of the GOP or decoding and making the effective data length until decoding is successful. it can. The same applies to compressed audio.

  An error having an effective data length of 0 is regarded as an error, but a special definition may be added to XML as in the Recording attribute of FIG. When the Recording attribute is true, it means that the clip is still being recorded, and indicates that the clip being recorded after restarting is in an abnormal state. In addition, Clip1. The file name being recorded may be different from the file name after the completion of recording, such as xml ~, and it may be identified that the file is being created as part of the file name.

  Although the clip metadata file is provided as the clip metadata, the clip information may be recorded in the header or footer of the essence file, not in a separate file. The operation of the abnormal clip detection means 240 in that case is the same as the detection method, except that the XML file or the MXF file is viewed.

In this embodiment, the clip metadata is XML, but other formats may be used.
(Embodiment 2)
In the first embodiment, when the clip is restored, the number of frames of the clip is determined in accordance with the number of valid frames of the audio data, so that video data for 4 frames is not referred to as a clip even though it can be played back. The recorded video data is wasted. In general, in a recording method in which data is written for each specific byte, audio data can store more time data in the specific byte. Therefore, the video data recorded in the restoration method of the first embodiment tends to be cut off. is there. In the present embodiment, a method of repairing a clip without waste by multiplexing and recording audio on a video will be described.

  As shown in FIG. 4, the DIF shown in this embodiment can multiplex and record audio on video. However, considering after-recording, there is no guarantee that the audio multiplexed in the video always matches the audio file. Therefore, an in-video audio data identification flag called ValidAudioFlag as shown in FIG. 13 is added to the metadata. When ValidAudioFlag is true, it indicates that the audio data multiplexed and recorded on the audio file and the video is the same data. This information is recorded in the clip metadata by the clip metadata creating means 221 at the time of recording.

  FIG. 14 shows a method for correcting a clip when the above-mentioned failure occurs when ValidAudioFlag is true. When the valid data adjustment unit 246 detects that the audio file is shorter than the video file, such as 157 frames for the video file and 153 frames for the audio file, as in the first embodiment, the ValidAudioFlag in the clip metadata is detected. Check the truth of. If ValidAudioFlag is true, the clip recognizes that recording has ended due to a failure, and the valid data adjustment unit 246 instructs the audio data extraction unit 270 to extract the deficient data from the video data. To do. The audio data extracting means 270 extracts the deficient 4 frames of audio data from the video data from the 154th frame to the 157th frame, and adds it to the back of the 153th frame of the audio data. The essence file metadata repairing unit 245 rewrites the portion related to the effective data length in the header and footer of the audio file to 157 frames. The valid data adjustment unit 246 records the clip length of 157 frames including the data extracted from the video in the clip metadata, and completes the restoration.

  By recording and restoring by the above method, it is possible to restore a clip without wasting video data. In addition, since the range of audio data extraction from video data can be narrowed compared to extracting all audio data from video data, the restoration process can be completed at high speed.

  The present invention can be applied to a photographing apparatus or the like that records data photographed on a random access medium such that an editing operation is assumed as a post process.

FAT format explanation diagram Directory entry illustration The figure which shows the example of the clip metadata of Embodiment 1 The figure which shows the structure of the essence file of Embodiment 1,2. Configuration diagram of photographing apparatus according to Embodiment 1 Directory configuration diagram of imaging apparatus according to Embodiment 1 Explanatory drawing of abnormal state of clip metadata in Embodiment 1 Configuration diagram of clip restoration according to Embodiment 1 The figure which shows the format confirmation means of Embodiment 1. Illustration of UMID The figure which shows another example 1 of the abnormal state of the clip metadata of Embodiment 1 The figure which shows another example 2 of the abnormal state of the clip metadata of Embodiment 1 The figure which shows the example of the audio | voice multiplexed information of Embodiment 2. Configuration diagram of clip restoration according to Embodiment 2

Explanation of symbols

100 Physical Drive 101 Logical Drive 102 Master Boot Record 103 Partition Boot Sector 104 File Allocation Table # 1
104b File allocation table # 2
105 Root directory entry 106 User data 107 Directory entry 108 First cluster number 201 EssenceList
202 ClipMetadata
210 File header 211 File body 212 File footer 213 Video file 214 Audio file 1
215 Audio file 2
216 Audio data A multiplexed in video file
217 Audio data B multiplexed in video file
220 File storage location management means 221 Clip metadata creation means 222 Video file creation means 223 Audio file creation means 224 Video input means 225 Audio input means 226 Recording means 227 Recording medium 240 Abnormal clip detection means 241 File format detection means 242 File format confirmation Means 243 Video reproduction device 244 Effective data length detection device 245 Essence file metadata restoration means 246 Effective data length adjustment means 247 Final frame metadata extraction means 250 Basic UMID
251 Source Pack 252 Universal Label 253 Length (Valid Data Length)
254 instance number
255 Material No
256 time 257 place 258 country 259 organization 260 user 270 voice data extraction means 280 format selection means 281 image display section

Claims (6)

A video imaging device that records one or more video and audio data as files on a recording medium,
File format detection means for detecting the file format of the file;
Effective data length detection means for detecting the length of data that has been recorded out of the data included in the file based on the format detected by the file format detection means;
A video photographing apparatus comprising: metadata restoration means for restoring metadata of the file based on the valid data length detected by the valid data length detection means. One or more video and audio data are each recorded on a recording medium as a file, and a plurality of recorded files are handled as one unit (clip),
An effective data length adjusting unit that calculates the number of effective frames of the clip from the effective data length of each of the video and audio data detected by the effective data length detecting unit and records the calculated number of frames in the metadata of the video and audio data. Item 2. The video photographing device according to Item 1. One or more video and audio data are each recorded as a file on a recording medium, and a clip metadata file that is information about the video and audio data is recorded as a file, and a plurality of recorded files are recorded as a unit (clip) As a video shooting device,
2. An effective data length adjusting unit for calculating the number of effective frames of the clip from the effective data lengths of the video and audio data detected by the effective data length detecting unit and recording the calculated number in the clip metadata file. Video shooting device. 4. The clip according to claim 2, wherein the clip retains normal recording success / failure information indicating whether or not the recording has been normally completed, and comprises an abnormal clip detection means for detecting an abnormal clip based on the normal recording success / failure information. The video imaging device according to any one of the above. Format confirmation means for confirming to the user the file format detected by the file format detection means;
2. The video photographing apparatus according to claim 1, further comprising video reproducing means for reproducing video and audio data in a format specified by the user at the time of confirmation. Audio extraction means for extracting audio data from a video file in which the same audio data as the audio data recorded as a file is multiplexed;
The effective data length adjustment unit refers to information indicating that the same audio data as the audio data recorded as a file is multiplexed in the video data, and instructs the audio extraction unit to extract audio from the video file. The video imaging device according to claim 2.

Images