JP2003283966A - Moving picture data digest information preparing device, method therefor, moving picture data recording/ reproducing device, and moving picture data recording/ reproducing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device and method which can prepare an digest image that enables a user to grasp especially important contents in video contents in a short time and which can prepare the digest image easily and simply. <P>SOLUTION: A moving picture encoder 20 encodes moving picture data; detects a color feature, a motion feature, a shape feature and so on; and stores the detected features in a moving picture feature database 60. On the basis of the data from the encoder 20 and the features stored in the feature database 60, an identical scene judger 70 detects identical scenes. A time code for the identical scenes is recorded in a display scene information database 80. A display scene selector 90, on the basis of the time code for the identical scenes, reads out moving picture data about the identical scene from an encoded moving picture database 50, a moving picture decoder 100 decodes the encoded data, and displays it on a display 110. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a video recording / reproducing device, and more particularly to a device for automatically generating and browsing an outline of video content.

[0002]

2. Description of the Related Art Conventionally, in a device for recording and reproducing video content, prior to viewing the video content,
It is known that a user can select a video content to be viewed based on the content by viewing a summary of the content.

Here, in generating the outline of the video contents, for example, there is a method in which a single frame or a group of frames is extracted as a scene in a certain time unit, and each scene is assembled into an outline. In addition, JP-A-6-2
In the index display device described in Japanese Patent No. 53255, one or more index images per scene are extracted from a video including a plurality of scenes, and the index images are used to show the relationship between indexes included in the scenes. It is displayed according to the form of expression.

[0004]

However, in the summary generated as described above, since the importance of each scene is not judged, the user can grasp the important contents in the contents in a short time. I can't. Also,
In the case of the index display device described in JP-A-6-253255, since the index image is only displayed, it is difficult for the user to know what is particularly important among the video contents. That is, since the index image is displayed for each scene, it is difficult to know the contents of high importance. Further, this Japanese Patent Laid-Open No. 6-25
In the case of the index display device described in No. 3255,
Complex content analysis must be performed on moving image content.

Therefore, the present invention is to provide a device and a method capable of creating a summary image capable of grasping particularly important contents in a video content in a short time and easily creating the summary image. It is intended to provide.

[0006]

The present invention was created to solve the above problems. First, a moving image data summary information creating apparatus for creating summary information of moving image data. Therefore, a feature amount detecting means for sequentially detecting feature amounts for each frame forming the moving image data, and a scene for dividing the moving image data into a plurality of scenes based on the feature amounts detected by the feature amount detecting means And a dividing unit, and an identical scene detecting unit that detects the identical scene by comparing the characteristic amount of a certain scene with the characteristic amount of one or a plurality of scenes located at a position before the scene. Characterize.

In the moving image data summary information creating apparatus having the first structure, the characteristic amount detecting means sequentially detects the characteristic amount for each frame constituting the moving image data. Then, the scene dividing unit divides the moving image data into a plurality of scenes based on the characteristic amount detected by the characteristic amount detecting unit, and the same scene detecting unit divides the moving image data into the characteristic amount in a certain scene and the scene before the scene. The same scene is detected by comparing with the feature amount in one or a plurality of scenes at the position. Therefore, the content of the moving image data can be known by browsing the detected same scene as a summary image. In particular, since the contents of the same scene can be viewed as a summary image, it is possible to grasp the important contents in the moving image data. In addition, in particular, when detecting the same scene, since the detection is performed based on the feature amount in each scene, it is possible to appropriately detect the same scene,
Also, the same scene can be easily detected.

Secondly, in the first configuration, the feature amount detected by the feature amount detecting means is any one of a color feature amount, a motion feature amount, and a shape feature amount, or any one of them. It is a combination of

Thirdly, there is provided a moving picture data recording / reproducing apparatus for recording / reproducing moving picture data, which comprises coding means for coding the moving picture data, and data coded by the coding means. And a feature amount detecting unit for sequentially detecting a feature amount for each frame forming the moving image data based on the data calculated by the encoding unit, and the feature amount detecting unit. A scene dividing unit that divides the moving image data into a plurality of scenes based on the detected feature amount, a feature amount in a certain scene, and a feature amount in one or a plurality of scenes at positions before the scene. An identical scene detection unit that detects the identical scene by comparison and outputs identification information that is information that identifies at least one of the detected identical scenes; Based on the specific information storage unit that stores the specific information output from the one scene detection unit and the specific information stored in the specific information storage unit, a predetermined image is stored in the moving image data stored in the moving image data storage unit. And a summary image reproducing means for reproducing the summary image composed of the predetermined moving image data by reading the moving image data.

In the third structure, the encoding means encodes the moving image data, and the encoded data is stored in the encoded data storage means. Further, the feature amount detecting means, based on the data calculated by the encoding means,
The feature amount is sequentially detected for each frame forming the moving image data. Then, the scene dividing unit divides the moving image data into a plurality of scenes based on the characteristic amount detected by the characteristic amount detecting unit, and the same scene detecting unit divides the moving image data into a characteristic amount in a certain scene and a scene before the scene. The same scene is detected by comparing with the feature amount of one or a plurality of scenes at the position, and specific information that is information for specifying at least one of the detected same scenes is output, and the specified scene is output. The information is stored in the specific information storage means. Then, the summary image reproduction means reads out predetermined moving image data from the moving image data stored in the moving image data storage means on the basis of the specific information stored in the specific information storage means, and outputs the predetermined moving image. Play a summary image of data.

Therefore, since the detected same scene can be viewed as a summary image, important contents in the moving image data can be grasped. Further, in particular, when the same scene is detected, the detection is performed based on the feature amount in each scene, so that the same scene can be appropriately detected, and the same scene can be easily detected.

Fourthly, in the third configuration, the feature quantity detected by the feature quantity detecting means is any one of a color feature quantity, a motion feature quantity, and a shape feature quantity, or any one of them. It is a combination of

Fifthly, in the third or fourth configuration, the encoding means performs a DCT conversion on the data of the frame forming the moving image data and outputs the DCT portion. A quantizer that quantizes and outputs the data output from the DCT unit; and a variable-length encoder that variable-length-encodes the data output from the quantizer,
A local decoding unit that performs a local decoding process on the data output from the quantizing unit, and a motion vector is detected based on the data output from the local decoding unit and the frame data input to the encoding unit. Then
A motion compensation unit that performs motion compensation on the data of the frame based on the detected motion vector and outputs the motion compensation data to the DCT unit, and the feature amount detection unit outputs the output from the local decoding unit. The color feature quantity and the shape feature quantity are detected based on the above, and the motion compensation section detects the motion feature quantity based on the motion vector detected by the motion compensation section.

A sixth aspect is a moving image data recording / reproducing apparatus for recording / reproducing moving image data, wherein an encoding means for encoding the moving image data and a moving image encoded by the encoding means. Encoded data storage means for storing image data, decoding means for decoding data stored in the encoded data storage means, decoded data storage means for storing data decoded by the decoding means, and decoding means A plurality of moving image data based on the feature amount detected by the feature amount detecting means for sequentially detecting the feature amount for each frame forming the moving image data based on the data calculated in By comparing the scene dividing means for dividing the scene into a scene, a feature amount in a certain scene, and a feature amount in one or a plurality of scenes at positions before the scene, Identical scene detecting means for detecting one scene and outputting specific information which is information for identifying at least one of the detected identical scenes, and specific information for storing the specific information output from the identical scene detecting means It is characterized by comprising a storage means and a summary image reproduction means for reproducing predetermined data from the data stored in the decoded data storage means based on the specific information stored in the specific information storage means.

In the sixth configuration, the encoding means encodes the moving image data, and the encoded data is stored in the encoded data storage means. Further, the decoding means decodes the data stored in the encoded data storage means and stores it in the decoded data storage means. Further, the feature amount detecting means sequentially detects the feature amount for each frame forming the moving image data based on the data calculated by the decoding means. Then, the scene dividing unit divides the moving image data into a plurality of scenes based on the characteristic amount detected by the characteristic amount detecting unit, and the same scene detecting unit divides the moving image data into a characteristic amount in a certain scene and a scene before the scene. The same scene is detected by comparing with the feature amount of one or a plurality of scenes at the position, and specific information that is information for specifying at least one of the detected same scenes is output, and the specified scene is output. The information is stored in the specific information storage means. Then, the summary image reproduction means reproduces predetermined data from the moving image data stored in the decoded data storage means based on the specific information stored in the specific information storage means.

Therefore, since the detected same scene can be viewed as a summary image, important contents in the moving image data can be grasped. Further, in particular, when the same scene is detected, the detection is performed based on the feature amount in each scene, so that the same scene can be appropriately detected, and the same scene can be easily detected.

Seventhly, in the sixth configuration, the feature quantity detected by the feature quantity detecting means is any one of a color feature quantity, a motion feature quantity, and a shape feature quantity, or any one of them. It is a combination of

Eighthly, in the sixth or seventh configuration, the decoding means performs variable length decoding of the encoded data to output the decoded data, and the motion vector of the frame to be decoded is A variable-length decoding unit that outputs data, an inverse quantization unit that inversely quantizes and outputs the decoded data output from the variable-length decoding unit, and a data output from the inverse quantization unit A DCT inverse transform unit for performing DCT inverse transform, and outputting the result. The feature amount detecting means detects a color feature amount and a shape feature amount based on the data output from the DCT inverse transform unit. The feature amount detecting means is characterized by detecting a motion feature amount from the data output from the DCT inverse transform unit or the motion vector data output from the variable length decoding unit.

Ninthly, in any one of the third to eighth constructions, the scene dividing means has a feature amount in a target frame which is a target of feature amount detection, and the target frame. It is characterized by dividing each scene while detecting the boundary position with the previous scene by comparing with the feature quantity in one or a plurality of frames located before the scene.

Tenth, in any one of the third to ninth configurations, when the scene dividing means detects a scene before the target frame, the same scene detecting means is provided. Is a histogram of the feature amount in the detected scene and 1 in front of the scene.
Alternatively, it is characterized in that whether or not the scenes are the same is determined by comparing with a histogram of the feature amount in each scene in a plurality of scenes. Therefore, it is possible to appropriately detect the same scene.

Eleventh, in any one of the third to tenth configurations, the summary image reproduction means reproduces only the scene specified by the specification information.

Twelfthly, in any one of the third to eleventh configurations, the moving image data recording / reproducing apparatus is further provided with reproduction start position information provided in accordance with the genre of the moving image data. A reproduction start position information, which is information about a reproduction start position in the moving image data, in which at least one reproduction start position information stores reproduction start position information defined in association with the specific information. Storage means and moving image data reproducing means for reproducing the moving image data stored in the encoded data storage means or the decoded data storage means, wherein the reproduction start position information stored in the reproduction start position information is reproduced. Moving image data reproducing means for reproducing the moving image data from the reproduction start position indicated by the reproduction start position information corresponding to the genre of the target moving image data, Characterized in that it has.

As a result, the reproduction start position can be changed depending on the genre of the moving image data, so that proper viewing can be performed according to the genre.

Thirteenth, in any one of the third to twelfth configurations, the same scene detecting means specifies the specific information which is information for specifying each scene in the detected same scene. A moving image data recording / reproducing device for reproducing the moving image data stored in the encoded data storage unit or the decoded data storage unit, the specific information being stored in a storage unit; It is characterized in that only one scene is reproduced for the same scene based on the specific information stored in the storage means. Therefore, since only one scene is reproduced for the same scene, all the scenes can be viewed in a short time.

The fourteenth is a moving picture data summary information creating method for creating summary information of moving picture data,
A characteristic amount detecting step of sequentially detecting a characteristic amount for each frame forming the moving image data, and a scene dividing step of dividing the moving image data into a plurality of scenes based on the characteristic amount detected by the characteristic amount detecting step. , The same scene detection step of detecting the same scene by comparing the feature quantity in a certain scene with the feature quantity in one or a plurality of scenes located at a position prior to the scene. .

Therefore, the contents of the moving image data can be known by browsing the detected same scene as a summary image. In particular, since the contents of the same scene can be viewed as a summary image, it is possible to grasp the important contents in the moving image data. In addition, in particular, when detecting the same scene, since the detection is performed based on the feature amount in each scene, it is possible to appropriately detect the same scene,
Also, the same scene can be easily detected.

Fifteenth, moving image data is recorded and
A moving image data recording / reproducing method for reproducing, comprising an encoding step of encoding moving image data, an encoded data storing step of storing the data encoded by the encoding means in a predetermined storage device, A feature amount detecting step of sequentially detecting a feature amount for each frame forming the moving image data based on the data calculated in the encoding step, and a moving image based on the feature amount detected by the feature amount detecting step. A scene dividing step of dividing data into a plurality of scenes, a feature amount in a certain scene, and a position 1 before the scene.
Alternatively, the same scene detection step of detecting the same scene by comparing the feature amounts of a plurality of scenes and outputting specific information that is information for specifying at least one of the detected same scenes, A specific information storage step of storing the specific information output in the scene detection step in a predetermined storage device, and moving image data stored in the moving image data storage means based on the specific information stored in the specific information storage step. From the predetermined moving image data and reproduces a summary image composed of the predetermined moving image data.

Therefore, since the detected same scene can be browsed as a summary image, important contents in the moving image data can be grasped. Further, in particular, when the same scene is detected, the detection is performed based on the feature amount in each scene, so that the same scene can be appropriately detected, and the same scene can be easily detected.

Sixteenth, recording and recording moving image data.
A moving image data recording / reproducing method for reproducing, comprising an encoding step of encoding moving image data, and an encoded data storing step of storing the moving image data encoded by the encoding means in a predetermined storage device. A decoding step of decoding the data stored in the moving image data storage step, a decoded data storage step of storing the data decoded in the decoding step in a predetermined storage device, and a data calculated in the decoding step. A feature amount detecting step of sequentially detecting a feature amount for each frame constituting the moving image data, and a scene in which the moving image data is divided into a plurality of scenes based on the feature amount detected by the feature amount detecting step. The division process, the feature amount in a certain scene, and the position 1 before the scene
Alternatively, the same scene detection step of detecting the same scene by comparing the feature amounts of a plurality of scenes and outputting specific information that is information for specifying at least one of the detected same scenes, Based on the specific information storage step of storing the specific information output from the scene detection means in a predetermined storage means, and the specific information stored in the specific information storage step, the predetermined data is stored in the decoded data storage step. And a summary image reproducing step of reproducing the data of.

Therefore, since the detected same scene can be viewed as a summary image, it is possible to grasp the important contents in the moving image data. Further, in particular, when the same scene is detected, the detection is performed based on the feature amount in each scene, so that the same scene can be appropriately detected, and the same scene can be easily detected.

[0031]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described with reference to the drawings. A moving image data recording / reproducing apparatus and a moving image data recording / reproducing apparatus A1 as a moving image data summary information creating apparatus according to the first embodiment of the present invention are
The frame memory 10 is configured as shown in FIG.
A moving image encoding unit 20, an encoded moving image database (encoded data storage unit) 50, a moving image feature amount database 60, the same scene determination unit 70, and a display scene information database (specific information storage unit). 80 and operation unit 8
5, a display scene selection unit 90, and a moving image decoding unit 100.
And a display unit 110.

Here, the moving picture coding unit 20 MPEG-codes the moving picture data read from the frame memory 10 and records it in the coded moving picture database 50, and at the same time, the moving picture data read from the frame memory 10. The moving image feature amount of is calculated and stored in the moving image feature amount database 60, and is output to the same scene determination unit 70.

The structure of the moving picture coding unit 20 will be described in more detail. As shown in FIG. 2, the moving picture coding unit 20 includes a DCT unit 22, a Q unit 24, a VLC unit (variable length). It has an encoding unit 26, an ME / MC unit (motion compensation unit) 28, a local decoding unit 30, and a feature amount calculation unit (feature amount detection means) 40.

Here, the DCT unit 22 performs DCT on the moving image data read from the frame memory 10 or on the data sent from the ME / MC unit (motion compensation unit) 28.
It is a conversion. The Q section 24 quantizes the moving image data DCT-transformed by the DCT section 22. The VLC unit 26 also performs variable-length coding on the moving image data quantized by the Q unit 24. The ME / MC unit 28 detects a motion vector and performs motion compensation based on the detected motion vector. Here, the DCT unit 22, the Q unit (quantization unit) 24, and the VLC unit (variable length coding unit) 26 function as the coding means.

The local decoding section 30 has an IQ section 32, an IDCT section 34, an MC section 36, and a local decoded image memory section 38.

Here, the IQ section 32 inversely quantizes the quantized moving image data from the Q section 24. Further, the IDCT unit 34 performs an inverse DCT transform on the moving image data dequantized by the IQ unit 32. The MC unit 36 also performs motion compensation based on the motion vector. The local decoded image memory unit 38 also stores the data subjected to the inverse DCT conversion by the IDCT unit 34 and the motion-compensated data by the MC unit 36.

The feature quantity computing section 40 has a motion feature quantity computing section 42, a color feature quantity computing section 44, and a shape feature quantity computing section 46.

Here, the motion characteristic amount calculation unit 42 calculates a motion characteristic amount, and specifically calculates a motion descriptor in MPEG-7. That is, the ME / MC unit 2
Based on data from 8, MotionActiv
Compute a motion descriptor such as ty.

The color feature amount calculator 44 calculates a color feature amount, specifically, a color descriptor in MPEG-7. That is, the data for one frame stored in the locally decoded image memory unit 38 is vertically
It is divided into a total of 64 regions that are equally divided into 8 equal parts in the horizontal direction, the average value of pixel values is calculated in each region, and DCT conversion is performed on the total 64 average values to calculate the calculated DCT Among the numerical values, the coefficient value of a predetermined upper number is used as a color descriptor.

The shape feature amount calculator 46 calculates the shape feature amount, and specifically, ISO / IEC 1
5938-3 MPEG-7 Visual (hereinafter referred to as M
The shape descriptor in PEG-7) is calculated.

The feature amount calculator 40 stores each calculated feature amount in the moving image feature amount database 60.

Further, the encoded moving image database 50 is
The storage device stores moving image data encoded by the moving image encoding unit 20.

Further, the moving image feature amount database 60 is
It is a storage device that stores the feature amount calculated by the feature amount calculation unit 40, and as shown in FIG.
The color feature amount, the motion feature amount, and the shape feature amount are stored.

The same-scene determination unit 70 determines the same scene in the moving image data. Specifically, the same-scene determination unit 70 compares the feature amount of the current frame with the feature amount of the immediately preceding frame to generate a moving image. The image data is divided for each scene, and it is detected whether the same scene exists by comparing the immediately preceding scene with the scene before the immediately preceding scene,
When the same scene is detected, information for specifying the scene is stored in the display scene information database 80. The same scene determination unit 70 functions as the same scene detection unit as the scene division unit.

Specifically, the same-scene determination unit 70 has a same-scene determination table as shown in FIG. 4, and the same-scene determination table has a start time code and a The feature quantity histogram can be stored. Then, while detecting the break of the scene, the data is stored in the same scene determination table, and it is determined whether or not they are the same scene based on the data stored in the table. Details will be described later.

Further, the display scene information database 80
Is a storage device that stores data for specifying the determined scene of the same scene. Specifically, the configuration is as shown in FIG. 5, and the time code, that is, the time information of the start time and the time information of the end time of the scene is stored for each same scene.

The display scene selection unit 90 cuts out predetermined moving image data from the encoded moving image database 50 based on the data stored in the display scene information database 80 and sends it to the moving image decoding unit 100. .

The moving picture decoding section 100 is for decoding the moving picture data sent from the display scene selecting section 90. Specifically, the moving picture decoding section 24 shown in FIG.
It has the same configuration as 0. That is, the moving image data sent from the display scene selection unit 90 is decoded by performing variable length decoding, inverse quantization, I-Scan, IDCT (inverse DCT conversion).

Further, the display unit 110 includes the moving picture decoding unit 1.
The moving picture data decoded by 00 is displayed.

It should be noted that, in the above-mentioned configuration, the feature amount calculation unit 40 is described as being provided in the moving image coding unit 20, but the feature amount calculation unit 40 is provided separately from the moving image coding unit 20. May be.

The moving image data recording / reproducing apparatus A1
Each of the above-described units may be configured as a device having each function, or a part of the configurations, for example, the moving image encoding unit 20, the same scene determination unit 70, the display scene selection unit 90, and the moving image decoding unit. The conversion unit 100 may include a program for executing the process of each unit and a CPU that executes the process based on the program.

A moving image data recording / reproducing apparatus A1 having the above structure.
The operation of will be described. First, the operation of detecting the same scene will be described with reference to the flowchart shown in FIG. In the following description of the operation, for convenience of description, the feature amount calculation unit 40 calculates only the color feature amount. That is, the color feature amount calculation unit 44 calculates the color feature amount, but the motion feature amount calculation unit 42 and the shape feature amount calculation unit 46 do not perform the calculation.

When the moving image data is input to the frame memory 10, the moving image encoding unit 20 encodes the moving image data (encoding step). That is, encoding is performed by using the frame data stored in the frame memory 10 as a target frame to be processed. That is, I
In the case of intra image data that does not refer to another frame such as a picture, the image data from the frame memory 10 is input to the DCT unit 22, and the DCT unit 22 performs DCT conversion. After that, the Q section 24 performs quantization, the VLC section 26 performs variable length coding, and the variable length coding is stored in the coded moving image database 50. In the case of inter image data that refers to another frame such as a B picture, the ME / MC unit 28 moves based on the data stored in the local decoded image memory unit 38 and the data from the frame memory 10. The vector is calculated, the reference image is cut out from the local decoded image memory unit 38 according to the motion vector, and the difference from the data from the frame memory 10 is calculated. Similarly, the difference data is DCT-transformed by the DCT unit 22, quantized by the Q unit 24, and variable-length coded by the VLC unit 26. The coding process of moving image data as described above is similar to the conventional case. The step of storing the encoded data in the encoded moving image database 50 as described above corresponds to the encoded data storing step.

Further, the moving picture coding unit 20 calculates the moving picture feature amount for the target frame (feature amount calculating step).
(S10). That is, the characteristic amount calculation unit 40 calculates the characteristic amount. Here, it is assumed that only the color feature amount is calculated. That is, the color feature amount calculation unit 44 calculates the color feature amount based on the data from the local decoded image memory unit 38.
Specifically, a color descriptor in MPEG-7 is calculated.

Then, the calculated feature amount is sent to the same scene determination unit 70 and stored in the moving image feature amount database 60 (S11). That is, the feature amount is stored for each frame data. Note that, here, it is assumed that only the color feature amount is stored in the moving image feature amount database 60. In addition, in reality, each feature amount is stored for each of the Y signal, the Cb signal, and the Cr signal.

Then, the same scene determination unit 70 determines the moving image feature amount sent from the moving image encoding unit 20 and the moving image of the frame before the frame (that is, the target frame) for the moving image feature amount. The feature amounts are compared to determine whether the difference is within a predetermined threshold value (S12).
Here, the frame in front of the target frame is a plurality of continuous frames that are traced back from the frame immediately before the target frame, and the moving image feature amount is such that the moving image feature amount of each frame is averaged. The processed feature quantity is used. Note that the moving image feature amount of the frame before the target frame is read from the data stored in the moving image feature amount database 60. It should be noted that here, the color feature amounts are compared.

If it is within the predetermined threshold value, it is judged that the scene is still in the same scene, and the step S16 is performed.
On the other hand, if the threshold value is exceeded, it is regarded as another scene and the time code of the current frame is held as the start time code of the new scene (S13). In this step S13, when the start position of a new scene is detected, the immediately preceding scene is specified. Therefore, the same scene determination unit 70 calculates the histogram of the feature amount of the immediately preceding scene. , The same scene determination table (see FIG. 4). That is, the feature amount of the frame included in the immediately preceding scene is read from the moving image feature amount database 60 and the histogram thereof is calculated. Here, only the histogram for the color feature amount is created. As for the feature quantity of the frame included in the immediately preceding scene, each feature quantity up to the frame corresponding to the start time code of the immediately preceding scene is traced back from the frame immediately before the current frame. It may be read from the quantity database 60. The steps S12 and S13 correspond to the scene dividing step. In addition, the same scene determination unit 70 in this case functions as the scene dividing unit.

From step S13, the process proceeds to step S14, and the histogram of the feature amount in the immediately preceding scene (that is, the scene in which the frame preceding the first frame of the new scene detected in step S13 is the final frame of the scene) And the histograms of the feature amounts of all the scenes before the scene are compared with each other to determine whether there is a scene having the same histogram (S14). Here, the histogram of the color feature amount of the immediately preceding scene and the histogram of the color feature amounts of all the scenes before the scene are compared. This step S14 corresponds to the same scene detection step. Further, the same scene determination unit 70 in this case functions as the same scene detection means.

If there are scenes having the same histogram, the time code of the immediately preceding scene is stored in the display scene information database 80 (specific information storing step) (S15). That is, the information about the start time and the information about the end time of the immediately preceding scene are stored. For the start time, store the start time code of the scene in the same scene determination table,
As for the end time, the start time code of the scene next to the scene may be stored. The time code stored in the display scene information database 80 corresponds to the specific information. When the processing in step S15 is completed, the process proceeds to step S16.

On the other hand, if it is determined in step S14 that there is no scene having the same histogram, the process proceeds to step S16.

In step S16, the moving image encoding unit 20 determines whether or not there is a next frame. If there is a next frame, the process returns to step S10, while
If there is no next frame, the process ends.

In step S15, the moving image coded data of the scene indicated by the time code may be stored together with the time code for the same scene.

By performing the processing as described above, the time code for the same scene is accumulated in the display scene information database 80.

Next, the operation in reproduction will be described. First, the display of the summary will be described. When the user gives an instruction to browse the summary through the operation unit 85, the display scene selection unit 90 causes the display scene information database 80 to be displayed.
Read the data stored in. That is, the time code for the same scene is read. Then, the display scene selection unit 90 reads the moving image data corresponding to the time code from the encoded moving image database 50 based on the read time code, and the moving image decoding unit 100.
Send to. Then, the moving image decoding unit 100 decodes the moving image data and displays it on the display unit 110.

As described above, when the moving picture coded data of the scene indicated by the time code is also stored in the display scene information database 80 in addition to the time code for the same scene, Scene selection unit 90
Reads out the moving image coded data stored in the display scene information database 80 and sends it to the moving image decoding unit 100 as it is for display.

FIG. 7 shows an example in which the above operation is performed. That is, the moving image data 310 is the scene 31
Divided into multiple scenes from 1 to 319, scene 31
2 and the scene 314 are the same scene, and the example in which the scene 316 and the scene 318 are detected as the same scene is shown. In this case, the time codes of the scene 314 and the scene 318 are stored in the display scene information database 80.

When displaying the summary, the time codes of the scene 314 and the scene 318 are read out and the scene 314 and the scene 318 are displayed. That is, when browsing the displayed summary, only the same detected scene is displayed. The display of the summary as described above corresponds to the summary image reproducing step. Further, the display scene selection unit 90, the moving image decoding unit 100, and the display unit 110 at the time of displaying this summary function as the above-mentioned summary image reproducing means.

As described above, according to the moving image data recording / reproducing apparatus A1 of the present embodiment, since the detected same scene can be viewed as a summary image, important contents in the moving image data can be grasped. Further, in particular, when the same scene is detected, the detection is performed based on the feature amount in each scene, so that the same scene can be appropriately detected, and the same scene can be easily detected.

After the summary is displayed as described above, the following detailed display operation may be performed. That is, after the display of the summary is completed, the display unit 110 is queried as to whether or not more detailed browsing is to be performed.
When the user performs an operation for performing the detailed browsing, the following detailed display is performed. Note that the following detailed display may be automatically performed when the display of the summary is completed.

Then, the display scene selection unit 90 reads out the moving image data from the encoded moving image database 50 and sends it to the moving image decoding unit 100 for display. At this time, the moving image data is read according to the genre. Make the position different. For example, for moving image data of a certain genre, the start position of the moving image data is the read position (this is the first mode), and for moving image data of another certain genre, it is the start position of the first identical scene. (This is referred to as the second mode.) For moving image data of another certain genre, the last same scene in the same scene group consisting of the same scenes is defined as the start position of the first scene (this is the second position). 3 modes) and so on. The first mode is a case where the moving image data is viewed from the beginning, and for example, a drama or the like is suitable as the genre. Further, since the second mode starts from the same scene at the beginning, a sports program or the like is suitable, for example. Further, in the third mode, since the last same scene first appears in the same scene group consisting of the same scenes, it is possible to delay the start position most while satisfying at least the condition of not missing important scenes. Since you can do it, you can enjoy efficient viewing,
For example, it can be said that it is suitable for variety programs. That is, the correspondence between the genre of moving image data and the mode to be applied is predetermined, and the predetermined mode is applied according to the genre of the moving image data.

FIG. 8 shows a specific example of the above detailed display. In the moving image data 410, a scene 412 is displayed.
And scene 414 are the same scene, and scene 4
16 and the scene 418 are the same scenes.

Then, when the mode corresponding to the genre of the moving image data 410 is the first mode, the display starts from the start position of the moving image data. That is, it is similar to the case of displaying moving image data normally. In the case of the second mode, the display is started from the start position of the scene 412 that first appears. In the third mode, the last identical scene appears first in the identical scene group consisting of the scenes 412 and 414 and the identical scene group consisting of the scenes 416 and 418. Since it is 414, display is started from the start position of this scene 414. If the scene 414 is after the scene 418, the display starts from the scene 418.

In order to perform the detailed browsing operation as described above, it is necessary to store all the time codes of the same scene in the display scene information database 80 or the like. In addition, a storage unit that stores the data about the start position for each genre is also provided (in this case,
The storage section corresponds to the reproduction start position storage means). For example, the information on each start position in the first mode to the third mode is stored in association with the information on the genre. Further, the genre of the target moving image data is acquired from the EPG data as shown in FIG. That is, data relating to the genre of the program is added to the moving image data input to the frame memory 10 in advance, and the information relating to the genre is extracted and stored. Then, the display scene selection unit 90 determines the start position according to the information regarding the genre, and reads the data of the encoded moving image database 50 from the start position. In this way, the display scene selection unit 90, the moving image decoding unit 100, and the display unit 110 in the case of changing the start position for each genre function as the moving image data reproducing unit.

As another method of the detailed browsing, a method of displaying only one scene for the same scene and efficiently browsing all of the moving image data can be considered as follows.

For example, in the same scene group consisting of the same scenes, only the first scene is displayed and the same scenes thereafter are not displayed. In other words, all time codes for the same scene are stored, and when displaying moving image data, the display starts from the start position of the moving image data, but the second and subsequent ones in the same scene are applicable. The display is omitted without reading the moving image data of the time code. For example, in the example of FIG.
As shown in (b), the display of scenes 414 and 418 is omitted.

In the above description, only the color feature amount is calculated as the feature amount, but if all the color feature amount, the motion feature amount and the shape feature amount are calculated, the above step S10 is performed. Of course, each characteristic amount is calculated, and in step S11, each characteristic amount is stored in the moving image characteristic amount database 60.

Further, in step S12, the criteria for the case of shifting to step S16 and the case of shifting to step S13 are set, and the judgment is made based on the criteria. For example, if all the differences for each feature amount are within the threshold value, the process proceeds to step S16. That is, when the difference is within the threshold for the color feature, the difference is within the threshold for the motion feature, and the difference is within the threshold for the shape feature, step S
If 16 is not within the threshold, the process proceeds to step S13. Of the three feature quantities,
If the predetermined number of feature quantities is within the threshold, step S16
If the number of feature quantities falling within the threshold value does not reach the predetermined number, the procedure may move to step S13.

Further, also in step S14, the criteria for the case of proceeding to step S16 and the case of proceeding to step S15 are set, and the determination is made based on the criteria. For example, if there is a scene in which the histograms are the same for each feature amount, the process proceeds to step S15. That is, comparing the immediately preceding scene with all the scenes before the immediately preceding scene, the histograms are the same for the color feature amount, the histograms are the same for the motion feature amount, and the shape feature amount is the same. If the histograms are the same, the process proceeds to step S15 and the time code of the immediately preceding scene is stored. Note that if the histograms are the same for a predetermined number of feature amounts of the three feature amounts, they may be regarded as the same scene and the time code of the immediately preceding scene may be stored.

Next, the second embodiment will be described. In the second embodiment, the moving image feature amount is calculated and the same scene is detected in the encoding in the first embodiment, whereas the moving image feature amount is calculated and the same scene is detected in the decoding. The difference is that the scene is detected.

The moving picture data recording / reproducing apparatus A2 of the second embodiment is configured as shown in FIG. 10, and has a frame memory 210 and a moving picture coding unit (coding means) 220.
An encoded moving image database (encoded data storage unit) 230, a moving image decoding unit (decoding unit) 240,
A feature amount calculation unit (feature amount detection means) 260, a moving image feature amount database 270, an identical scene determination unit 280,
Display scene information database (specific information storage means) 29
0, a display scene selection unit 300, a display scene memory (decoded data storage unit) 310, and a display unit 320.

Here, the moving picture coding unit 220 uses the moving picture data read from the frame memory 210 as MPEG.
It is encoded and recorded in the encoded moving image database 230. That is, it can be said that the configuration is obtained by removing the configuration of the feature amount calculation unit 40 from the moving image encoding unit 20 of the first embodiment.

Further, the encoded moving image database 230
Stores the moving image data encoded by the moving image encoding unit 220.

The moving picture decoding section 240 is for decoding the moving picture data stored in the coded moving picture database 230, and as shown in FIG. 11, a VLD section (variable length decoding section) 242. , IQ section (inverse quantization section) 244
, I-Scan unit 246, IDCT unit (DCT inverse transform unit) 248, frame memory 250, and MC unit 25.
2 and an adder 254.

Here, the VLD unit 242 is for performing variable length decoding on the encoded moving image data. Also, IQ
The unit 244 inversely quantizes the variable-length decoded moving image data. The I-Scan unit 246 zigzag-scans the dequantized moving image data, and the IDCT unit 248 performs an inverse DCT transform. Also,
The MC unit 252 reads the reference image from the frame memory 250 based on the motion vector sent from the VLD unit 242. It can be said that the configuration of the moving image decoding unit 240 itself is the same as the conventional configuration.

Further, the feature quantity computing section 260 computes a moving picture feature quantity, and as shown in FIG. 11, a color feature quantity computing section 262, a shape feature quantity computing section 264, and a motion feature quantity computing. Section 266.

Here, the color feature amount calculator 262 calculates the color feature amount, and specifically calculates the color descriptor in MPEG-7. That is, one frame is divided into a total of 64 pixels, each of which is divided into 8 equal parts in the vertical direction and 8 equal parts in the lateral direction, the average value of the pixel values is obtained for each of the set, and a 64 numerical value set consisting of the average values On the other hand, the two-dimensional DCT transformation based on the two-dimensional spatial position is performed, and several coefficient values in the upper order from the output coefficient values of this two-dimensional DCT transformation are used as color descriptors.

The shape feature amount calculator 264 calculates the shape feature amount, and specifically calculates the shape descriptor in MPEG-7.

The motion characteristic amount calculation unit 266 calculates the motion characteristic amount, and specifically calculates the motion descriptor in MPEG-7. That is, the motion feature amount calculation unit 266 has the ME unit 268 that detects the motion vector of the frame data itself based on the data from the frame memory 250, and the motion feature amount calculation unit 266 includes the intra image such as an I picture. In this case, the motion feature amount is calculated based on the motion vector detected by the ME unit 268,
On the other hand, in the case of an inter image such as a P picture or a B picture, the motion feature amount is calculated based on the motion vector sent from the VLD unit 242.

The characteristic amount calculation section 260 stores the calculated respective characteristic amounts in the moving image characteristic amount database 270.

The moving image feature amount database 270 stores the feature amount calculated by the feature amount calculating section 260, and stores the color feature amount, the motion feature amount, and the shape feature amount for each frame. It is configured to remember.
That is, the configuration of the moving image feature amount database 270 is similar to the configuration shown in FIG.

Further, the same-scene determination unit 280 determines the same scene in the moving image data, similarly to the same-scene determination unit 70 in the first embodiment, and specifically, the feature amount of the current frame. And the feature amount of the immediately preceding frame are compared, the moving image data is divided into each scene, and the immediately preceding scene and the scene before the immediately preceding scene are compared to determine whether the same scene exists. When the same scene is detected, the information for specifying the scene is stored in the display scene information database 290. The same scene determination unit 280
It functions as the same scene detection means as the scene division means.

Further, the display scene information database 290
Like the display scene information database 80, stores the data for specifying the determined scene of the same scene. That is, as shown in FIG. 5, a time code, that is, time information of the start time of the scene,
The time information of the end time can be stored for each same scene.

Further, the display scene selection unit 300 performs a process of erasing at least a part of the moving image data stored in the display scene memory 310 based on the data stored in the display scene information database 290. That is, the moving image data other than the same scene is deleted based on the data stored in the display scene information database 290.

Further, the display section 320 reads out and displays the moving image data stored in the display scene memory 310.

The moving image data recording / reproducing apparatus A2 is used.
Each of the above-mentioned units may be configured as a device having each function, or a part of the configuration, for example, the moving image encoding unit 220, the feature amount calculation unit 260, the same scene determination unit 280, and the display scene selection. Unit 300 and video decoding unit 24
0 may be configured by a program for executing the process of each unit and a CPU that executes the process based on the program.

A moving image data recording / reproducing apparatus A2 having the above structure.
The operation of will be described. First, the operation of detecting the same scene and creating a summary image will be described using the flowchart shown in FIG. Note that, in the following description of the operation, for convenience of explanation,
Only the color feature amount is calculated. That is, the color feature amount calculation unit 44 calculates the color feature amount, but the motion feature amount calculation unit 42 and the shape feature amount calculation unit 46 do not perform the calculation.

First, moving image data is stored in the frame memory 21.
When input to 0, the moving picture coding unit 220 codes the moving picture data (coding step). That is, the frame data stored in the frame memory 210 is encoded as the target frame to be processed. The encoded data is stored in the encoded moving image database 230 (encoded data storing step).

After that, the operation shifts to the operation of the flowchart of FIG. The processes of steps S21 to S25 in the flowchart of FIG. 12 are basically the same as steps S10 to S14 of the flowchart of FIG.

First, the moving picture decoding unit 240 decodes the coded moving picture data stored in the coded moving picture database 230 frame by frame, and stores the decoded moving picture data in the display scene memory 310. (Decoding step) (S20). That is, when the target frame is frame data that does not refer to another frame such as an I picture, the image data of the frame is subjected to variable length decoding by the VLD unit 242, inversely quantized by the IQ unit 244, and I-Scan is performed.
Inverse DC is performed by the IDCT unit 248 after scanning by the unit 246.
After the T conversion, it is stored in the frame memory 250. In the case of frame data that refers to another frame such as a B picture, the MC unit 252 cuts out the reference image from the frame memory 250 based on the motion vector and outputs the reference image to the addition unit 254. 248
The data is added to the data from and sent to the frame memory 250. The decoding process of moving image data as described above is similar to the conventional case.

Further, the characteristic amount calculation section 260 calculates the moving image characteristic amount for the target frame (feature amount detecting step).
(S21). Here, it is assumed that only the color feature amount is calculated. That is, the color feature amount calculation unit 262 calculates the color feature amount based on the data from the moving image decoding unit 240. Specifically, a color descriptor in MPEG-7 is calculated.

Then, the calculated feature amount is sent to the same scene determination unit 280 and stored in the moving image feature amount database 270 (S22). That is, the feature amount is stored for each frame data. Note that here, only the color feature amount is the moving image feature amount database 270.
Shall be stored in. Actually, the Y signal, C
Each feature amount is stored for each of the b signal and the Cr signal.

Then, the same scene determination unit 280 determines the moving image feature amount sent from the feature amount calculation unit 260 and the moving image feature of the frame before the frame (that is, the target frame) for the moving image feature amount. The amounts are compared to determine whether the difference is within a predetermined threshold value (S2
3). Here, the frame in front of the target frame is a plurality of continuous frames that are traced back from the frame immediately before the target frame, and the moving image feature amount is such that the moving image feature amount of each frame is averaged. The processed feature quantity is used. For the moving image feature amount of the frame before the target frame, the moving image feature amount database 2
It reads from the data stored in 70. It should be noted that here, the color feature amounts are compared.

If it is within the predetermined threshold value, it is judged that the scene is still in the same scene, and step S28 is performed.
On the other hand, if it exceeds the threshold value, it is regarded as another scene, and the time code of the current frame is held as the start time code of the new scene (S24). In step S24, when the start position of a new scene is detected, the immediately preceding scene is specified. Therefore, the same scene determination unit 280 calculates the histogram of the feature amount of the immediately preceding scene. , The same scene determination table (see FIG. 4). That is, the feature amount of the frame included in the immediately preceding scene is read from the moving image feature amount database 270 and the histogram thereof is calculated. Here, only the histogram for the color feature amount is created. As for the feature quantity of the frame included in the immediately preceding scene, each feature quantity up to the frame corresponding to the start time code of the immediately preceding scene is traced back from the frame immediately before the current frame. It may be read from the quantity database 270. The steps S23 and S24 correspond to the scene dividing step. In addition, the same scene determination unit 280 in this case
Functions as the scene dividing unit.

From step S24, the process proceeds to step S25, and the histogram of the characteristic amount in the immediately preceding scene (that is, the scene in which the frame preceding the first frame of the new scene detected in step S24 is the final frame of that scene) And the histograms of the characteristic amounts of all the scenes before the scene are compared with each other to determine whether or not there is a scene having the same histogram (S25). Here, the histogram of the color feature amount of the immediately preceding scene and the histogram of the color feature amounts of all the scenes before the scene are compared. This step S25 corresponds to the same scene detection step. Further, the same scene determination unit 280 in this case functions as the same scene detection means.

If there are scenes having the same histogram, the time code of the immediately preceding scene is stored in the display scene information database 290 (specific information storing step) (S26). That is, the information about the start time and the information about the end time of the immediately preceding scene are stored. For the start time, the start time code of the scene in the same scene determination table may be stored, and for the end time, the start time code of the next scene of the scene may be stored. The time code stored in the display scene information database 290 corresponds to the specific information. On the other hand, in step S25, if there is no scene having the same histogram, the process proceeds to step S28.

When the processing in step S26 is completed, the display scene selection unit 300 erases the predetermined moving image data (S27). That is, the moving image data from the time when the same scene is detected immediately before to the time when the same scene is detected is deleted from the display scene memory 310. That is, after the end time of the time code of the same scene stored before the time code of the same scene stored most recently in the display scene information database 290 and immediately before the most recently stored same scene. Delete the moving image data. When the processing of step S27 is completed, step S
Move to 28.

In step S28, the moving image decoding unit 240 determines whether or not there is a next frame. If there is a next frame, the process returns to step S20, while if there is no next frame, the process returns to step S20. Ends the process.

By performing the processing as described above, the moving image data of the same scene is accumulated in the display scene memory 310. This is the summary data. For example, in the example of FIG. 7, the moving image data of the scene 314 and the moving image data of the scene 318 are stored in the display scene memory 310.

Next, the operation in reproduction will be described. First, the display of the summary will be described. When the user gives an instruction to browse the summary using the operation unit 330, the display unit 320 reads out and displays the moving image data stored in the display scene memory 310. Since the moving image data of the same scene is stored in the display scene memory 310, the same scene can be viewed.

As described above, according to the moving image data recording / reproducing apparatus A2 in the present embodiment, the same detected scene can be viewed as a summary image, so that the important contents in the moving image data can be grasped. Further, in particular, when the same scene is detected, the detection is performed based on the feature amount in each scene, so that the same scene can be appropriately detected, and the same scene can be easily detected.

When performing the detailed display as in the first embodiment, the moving image data regarding the scene is deleted from the display scene memory 310 according to the genre of the moving image data. The moving image data for this detailed browsing is created in the display scene memory 310 separately from the summary data created in FIG.

For example, when the genre is a genre to which the first mode is applied, the moving image data is stored as it is without erasing a part thereof. In the second mode, moving image data until the same scene first appears is deleted. In the case of the third mode, the moving image data until the last identical scene first appears in the identical scene group is deleted.

In order to perform the detailed browsing operation as described above, it is necessary to store all the time codes of the same scene in the display scene information database 290 or the like. Further, a storage unit is also provided for storing the data on the start position for each genre (in this case, the storage unit corresponds to the reproduction start position storage means). For example, the information on each start position in the first mode to the third mode is stored in association with the information on the genre. Further, the genre of the target moving image data is acquired from the EPG data as shown in FIG. Then, the display scene selection unit 300 determines the start position according to the information regarding the genre, and deletes a part of the moving image data when necessary. In this way, the display scene selection unit 300, the display scene memory 310, and the display unit 320 when the start position is changed for each genre
Functions as the moving image data reproducing means.

As described above, by creating the data for detailed browsing, the data for detailed browsing can be displayed.

As another method of detailed browsing, when only one scene is displayed for the same scene, if the same scene appears for the second time and thereafter, those moving image data are deleted from the display scene memory 310. To do so.

In the description of FIG. 12 described above, only the color feature amount is calculated as the feature amount, but in the case of calculating all of the color feature amount, the motion feature amount and the shape feature amount, the above steps are performed. In S21, each feature amount is naturally calculated, and in step S22, each feature amount is stored in the moving image feature amount database 270.

Further, in step S23, the criteria for the case of proceeding to step S28 and the case of proceeding to step S24 are set, and the determination is made based on the criteria. For example, if all the differences are within the threshold value for each feature amount, the process proceeds to step S28. That is, when the difference is within the threshold for the color feature, the difference is within the threshold for the motion feature, and the difference is within the threshold for the shape feature, step S
If 28 is not within the threshold, the process proceeds to step S24. Of the three feature quantities,
If the predetermined number of feature quantities is within the threshold, step S28
If the number of feature quantities falling within the threshold value does not reach the predetermined number, the procedure may move to step S24.

Further, also in step S25, the criteria for the case of proceeding to step S28 and the case of proceeding to step S26 are set, and the determination is made based on the criteria. For example, if there is a scene in which the histograms are the same for each feature amount, the process proceeds to step S26. That is, comparing the immediately preceding scene with all the scenes before the immediately preceding scene, the histograms are the same for the color feature amount, the histograms are the same for the motion feature amount, and the shape feature amount is the same. If the histograms are the same, the process proceeds to step S26 and the time code of the immediately preceding scene is stored. Note that if the histograms are the same for a predetermined number of feature amounts of the three feature amounts, they may be regarded as the same scene and the time code of the immediately preceding scene may be stored.

[0119]

According to the moving picture data summary information creating device, the moving picture data recording / reproducing device, the moving picture data summary information creating method and the moving picture data recording / reproducing method according to the present invention,
Since the same detected scene can be viewed as a summary image, it is possible to grasp the important contents in the moving image data. Further, in particular, when the same scene is detected, the detection is performed based on the feature amount in each scene, so that the same scene can be appropriately detected, and the same scene can be easily detected.

Further, in particular, according to the moving image data recording / reproducing apparatus of the twelfth aspect, since the reproduction start position can be changed depending on the genre of the moving image data, the proper viewing is performed according to the genre. be able to.

Further, in particular, according to the moving picture data recording / reproducing apparatus of the thirteenth aspect, since only one scene is reproduced for the same scene, all the scenes can be viewed in a short time.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a moving image data recording / reproducing apparatus according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of a moving picture coding unit in the moving picture data recording / reproducing apparatus according to the first embodiment of the present invention.

FIG. 3 is an explanatory diagram showing a configuration of a moving image feature amount database.

FIG. 4 is an explanatory diagram showing a configuration of a same scene determination table.

FIG. 5 is an explanatory diagram showing a configuration of a display scene information database.

FIG. 6 is a flowchart for explaining the operation of the moving image data recording / reproducing apparatus according to the first embodiment of the present invention.

FIG. 7 is an explanatory diagram for explaining the operation of the moving image data recording / reproducing apparatus according to the embodiment of the present invention.

FIG. 8 is an explanatory diagram for explaining the operation of the moving image data recording / reproducing apparatus according to the embodiment of the present invention.

FIG. 9 is an explanatory diagram showing an example of EPG data.

FIG. 10 is a block diagram showing a configuration of a moving image data recording / reproducing apparatus according to a second embodiment of the present invention.

FIG. 11 is a block diagram showing the configurations of a moving image decoding unit and a feature amount calculation unit in a moving image data recording / reproducing apparatus according to the second embodiment of the present invention.

FIG. 12 is a flow chart for explaining the operation of the moving image data recording / reproducing apparatus according to the second embodiment of the present invention.

[Explanation of symbols]

A1, A2 moving image data recording / reproducing apparatus 10, 210 frame memory 20, 220 Moving picture coding unit 22 DCT section 24 Q section 26 VLC section 28 ME / MC section 30 Local decoding section 32 IQ section 34 IDCT Department 36 MC Department 38 Locally decoded image memory section 40, 260 Feature amount calculation unit 42 Motion feature amount calculator 44-color feature calculator 46 Shape feature calculation unit 50,230 Encoded moving image database 60,270 Moving image feature database 70, 280 Same scene determination unit 80,290 Display scene information database 90, 300 Display scene selection section 100, 240 Video decoding unit 110, 320 Display 310 Display scene memory

Front page continuation (51) Int.Cl. ⁷ Identification code FI theme code (reference) H04N 5/91 H04N 5/91 N 5D110 7/24 7/13 Z 11/04 G11B 27/02 B F term (reference) 5B075 ND12 NS01 UU40 5C052 AC08 CC01 DD04 5C053 FA06 FA14 FA23 GB06 GB19 GB22 GB26 GB37 HA29 JA22 KA05 LA06 5C057 AA07 DA03 EA03 EA07 ED07 ED10 PP05 MA05 MN04 GH01 MA01 GH08 MA11 GH05 MA11 SS12 TA00 TB04 TC12 TC24 TD01 TD03 TD05 TD08 TD10 TD12 TD15 UA02 UA05 UA33 5D110 AA13 AA29 BB01 DA12 DC05 DC16 EA06 EA08 EA12 FA02

Claims (16)

[Claims] 1. A moving picture data summary information creating apparatus for creating summary information of moving picture data, comprising: a feature quantity detecting means for sequentially detecting a feature quantity for each frame constituting the moving picture data; and the feature quantity. A scene dividing unit that divides the moving image data into a plurality of scenes based on the feature amount detected by the detecting unit, a feature amount in a certain scene, and a feature amount in one or a plurality of scenes at positions before the scene. And a same-scene detecting unit that detects the same scene by comparing the same with the moving-picture data summary information creating apparatus. 2. The feature amount detected by the feature amount detecting means is any one of a color feature amount, a motion feature amount, and a shape feature amount, or an arbitrary combination thereof. The moving image data summary information creation device described in. 3. A moving image data recording / reproducing apparatus for recording / reproducing moving image data, comprising: an encoding unit for encoding the moving image data; and an encoding unit for storing the data encoded by the encoding unit. A data storage unit, a feature amount detection unit that sequentially detects a feature amount for each frame forming the moving image data based on the data calculated by the encoding unit, and a feature detected by the feature amount detection unit A scene dividing unit that divides the moving image data into a plurality of scenes based on the amount, a feature amount in a certain scene, and a feature amount in one or a plurality of scenes at positions before the scene, The same scene detection means for detecting the same scene and outputting specific information for specifying at least one scene of the detected same scene; Specific information storage means for storing the specific information output from the output means, and a predetermined moving image from the moving image data stored in the moving image data storage means based on the specific information stored in the specific information storage means A moving image data recording / reproducing apparatus, comprising: a summary image reproducing means for reading data and reproducing a summary image composed of the predetermined moving image data. 4. The feature amount detected by the feature amount detecting means is any one of a color feature amount, a motion feature amount, and a shape feature amount, or an arbitrary combination thereof. The moving image data recording / reproducing apparatus as described in. 5. The encoding means applies DC to frame data forming moving image data.
A DCT unit that performs T-transform and outputs, a quantizer that quantizes and outputs the data output from the DCT unit, and a variable that performs variable-length coding on the data output from the quantizer A long coding unit, a local decoding unit that performs a local decoding process on the data output from the quantizing unit, and a data output from the local decoding unit and a frame data input to the coding means. A motion compensation unit that detects a motion vector based on the detected motion vector, performs motion compensation on the data of the frame based on the detected motion vector, and outputs the motion compensation data to the DCT unit. , The color feature amount and the shape feature amount are detected based on the output from the local decoding unit, and the motion compensation unit performs motion based on the motion vector detected by the motion compensation unit. Moving image data recording and reproducing apparatus according to claim 3 or 4, wherein detecting the symptom amount. 6. A moving image data recording / reproducing apparatus for recording / reproducing moving image data, comprising: encoding means for encoding moving image data; and moving image data encoded by the encoding means. Encoded data storage means, decoding means for decoding data stored in the encoded data storage means, decoded data storage means for storing data decoded by the decoding means, and data calculated by the decoding means Based on the feature quantity detecting means for sequentially detecting the feature quantity for each frame forming the moving picture data, and dividing the moving picture data into a plurality of scenes based on the feature quantity detected by the feature quantity detecting means. By comparing the scene dividing means, the feature amount in a certain scene, and the feature amount in one or a plurality of scenes located before the scene, the same scene An identical scene detection unit that outputs specific information that is information that is detected and that identifies at least one scene of the detected identical scenes; and a specific information storage unit that stores the specific information output from the identical scene detection unit , Based on the specific information stored in the specific information storage means,
A moving picture data recording / reproducing device, comprising: a digest image reproducing means for reproducing predetermined data from the data stored in the decoded data storing means. 7. The feature amount detected by the feature amount detecting means is any one of a color feature amount, a motion feature amount, and a shape feature amount, or an arbitrary combination thereof. The moving image data recording / reproducing apparatus as described in. 8. The variable-length decoding unit, wherein the decoding unit variable-length-decodes the encoded data and outputs the decoded data, and outputs the motion vector data for the frame to be decoded, and the variable-length decoding unit. An inverse quantization unit that performs inverse quantization on the decoded data output from the output unit; and a DCT inverse transformation unit that performs DCT inverse transformation on the data output from the inverse quantization unit and outputs The feature amount detection means detects a color feature amount and a shape feature amount based on the data output from the DCT inverse transformation unit, and the feature amount detection unit outputs the DCT inverse transformation unit. 8. The moving image data recording / reproducing apparatus according to claim 6, wherein the motion characteristic amount is detected from the generated data or the data of the motion vector output from the variable length decoding unit. 9. The scene dividing means compares a feature amount in a target frame, which is a feature target frame, with a feature amount in one or a plurality of frames located before the target frame. 9. The moving image data recording / reproducing apparatus according to claim 3, wherein each scene is divided while detecting a boundary position with respect to a previous scene. 10. When the scene dividing unit detects a scene before the target frame, the same scene detecting unit detects the histogram of the feature amount in the detected scene and the scene before the scene. 10. A feature amount histogram in each scene in one or a plurality of scenes is compared to determine whether or not the scenes are the same scene, 3 or 4 or 5 or 6 or 7 or 8 or 9. The moving image data recording / reproducing apparatus as described in. 11. The moving image according to claim 3, wherein the summary image reproducing means reproduces only a scene specified by the specifying information. Image data recording / reproducing device. 12. The moving image data recording / reproducing apparatus further includes a reproduction start position information provided according to a genre of the moving image data, which is information about a reproduction start position in the moving image data. In the position information, at least one reproduction start position information is stored in the reproduction start position storage means for storing the reproduction start position information defined in association with the specific information, and the encoded data storage means or the decoded data storage means. A moving image data reproducing means for reproducing moving image data, wherein the reproduction start position information stored in the reproduction start position information indicates the reproduction start position indicated by the reproduction start position information corresponding to the genre of the moving image data to be reproduced. And a moving image data reproducing means for reproducing the moving image data, according to claim 3 or 4 or 5 or 6 or 7 or 8. Moving image data recording and reproducing apparatus according to 9 or 10 or 11. 13. The same scene detection means stores specific information, which is information for specifying each scene in the detected same scene, in the specific information storage means, and the moving image data recording / reproducing apparatus further comprises: A moving image data reproducing means for reproducing the moving image data stored in the encoded data storing means or the decoded data storing means, and based on the specific information stored in the specific information storing means, only one scene for the same scene 13. The moving image data recording / reproducing device according to claim 3, 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12. 14. A moving picture data summary information creating method for creating summary information of moving picture data, comprising: a feature quantity detecting step of sequentially detecting a feature quantity for each frame constituting the moving picture data; A scene dividing step of dividing the moving image data into a plurality of scenes based on the characteristic amount detected by the detecting step, a characteristic amount in a certain scene, and a characteristic amount in one or a plurality of scenes at positions before the scene. And the same scene detecting step of detecting the same scene by comparing the same with the moving picture data summary information creating method. 15. A moving image data recording / reproducing method for recording / reproducing moving image data, comprising an encoding step of encoding moving image data, and a predetermined storage device for storing the data encoded by the encoding means. A coded data storing step for storing the coded data, a feature quantity detecting step for sequentially detecting a feature quantity for each frame forming the moving image data based on the data calculated in the coding step, and the feature quantity detecting step. A scene dividing step of dividing the moving image data into a plurality of scenes based on the feature amount detected by the feature amount, a feature amount in a certain scene, and a feature amount in one or a plurality of scenes at positions before the scene. The same scene detection step of detecting the same scene by comparing and outputting specific information that is information for specifying at least one scene among the detected same scenes. A specific information storage step of storing the specific information output in the same scene detection step in a predetermined storage device, and stored in the moving image data storage means based on the specific information stored in the specific information storage step. A moving image data recording / reproducing method, which comprises reading out predetermined moving image data from moving image data and reproducing a summarized image formed of the predetermined moving image data. 16. A moving image data recording / reproducing method for recording / reproducing moving image data, comprising: an encoding step of encoding the moving image data; and a predetermined step of providing the moving image data encoded by the encoding means. Encoded data storage step of storing in storage device, decoding step of decoding data stored in the moving image data storage step, and decoded data storage step of storing data decoded in the decoding step in a predetermined storage device A feature amount detecting step of sequentially detecting a feature amount for each frame constituting the moving image data based on the data calculated in the decoding step, and a feature amount detected by the feature amount detecting step, A scene dividing step of dividing moving image data into a plurality of scenes, a feature amount in a certain scene, and a feature amount in one or a plurality of scenes at a position before the scene. A same scene detection step of detecting the same scene by comparing the quantity and outputting specific information which is information for specifying at least one scene of the detected same scene; and the same scene detection means. Specific information storing step of storing the specified information in a predetermined storage means, and an abstract for reproducing the predetermined data from the data stored in the decoded data storing step based on the specific information stored in the specific information storing step An image reproducing process, and a moving image data recording / reproducing method.