JP2006101324A - Recording and reproducing apparatus and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a recording and reproducing apparatus capable of skipping a program until a scene wherein an unknown character plays in the case of applying the recording and reproducing apparatus to recording and reproduction of, e.g. a broadcast program and avoiding the occurrence of an event wherein many face images are coincident or no coincident face image exists as a result of face collation. <P>SOLUTION: The recording and reproducing apparatus for recording (video-recording) a broadcast program and reproducing the recorded program collates a face image of a character designated by a user with a face image of a character played in the program and reproduces the program when the coincidence is first obtained from the start of the collation after both the images are coincident. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a recording / reproducing apparatus and a recording / reproducing method for recording (recording) a television broadcast program and reproducing the recorded program, for example.

Conventionally, for example, in a recording / playback apparatus that records (records) a TV broadcast program and plays back the recorded program, the user wants to skip the program to a scene in which the character specified by the user appears. There was but it was impossible.
In such a recording / reproducing apparatus, an image search apparatus for searching for a character from a recorded program is known (for example, see Patent Document 1). This image search device performs special detection on a face from a video, and further identifies the detected face, thereby distinguishing and displaying the face of a person appearing in the video.
JP 2001-167110 A

The above-described conventional image search device is a method of searching by comparing the face image designated by the user with the face image of the character in the recorded video, so that the character who does not know the face is searched. Is difficult.
In addition, since a threshold value is used for the determination of face matching, in some cases, as a result of the matching, it may happen that a large number of face images are regarded as matching or that no face image is considered as matching. There's a problem.

  Therefore, for example, when the present invention is applied to recording / playback of a broadcast program, the program can be skipped to a scene in which a character with an unknown face appears, and a large number of face images are obtained as a result of face matching. It is an object of the present invention to provide a recording / reproducing apparatus and a recording / reproducing method capable of avoiding a situation in which there is no face image that is regarded as a match or that there is no face image regarded as a match.

  The recording / reproducing apparatus of the present invention includes a recording unit that records a video and information of a specific object that appears in the video, a storage unit that stores an image of the specific object in advance, and a specific unit recorded by the recording unit. A specific object image acquisition means for acquiring an image of a specific object designated by a user based on the information of the object from the storage means, an image of the specific object acquired by the specific object image acquisition means, and the recording means The matching unit that collates the image of a specific object appearing in the video and outputs the collation result, and if the collation by the collating unit results in a match between the two images, And a reproducing means for reproducing the video recorded by the recording means from the point in time when a match is obtained.

  The recording / reproducing apparatus of the present invention includes a recording unit that records information on a received broadcast program and characters appearing in the program, a storage unit that stores face images of the characters in advance, and the recording unit. The face image acquisition means for acquiring the face image of the character designated by the user based on the recorded character information from the storage means, the face image acquired by the face image acquisition means, and the recording means are recorded. The collation means for matching the face images of the characters appearing in the program and outputting the collation results, and if the collation by the collation means results in a match between both face images, the collation is started. And reproducing means for reproducing the program recorded by the recording means from the time when the first match is obtained.

  According to the recording / reproducing method of the present invention, a video and information of a specific object appearing in the video are respectively recorded, and an image of the specific object designated by the user based on the recorded specific object information is stored in advance. If the acquired image of the specific object and the image of the specific object appearing in the recorded video are collated, and a match between the two images is obtained, the collation is performed. The recorded video is reproduced from the time when the first match is obtained from the start.

  According to the present invention, for example, when applied to recording and playback of a broadcast program, a program can be skipped to a scene in which a character who does not know the face appears, and a large number of face images are obtained as a result of face matching. Can be provided, and a recording / reproducing apparatus and a recording / reproducing method capable of avoiding a situation in which there is no face image regarded as coincident can be provided.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
In the following description, a case where the present invention is applied to the recording / playback of a television broadcast program acquired from a television broadcast or a network as an example will be described, but the present invention is not limited to this, and for example, a photograph is taken with a television camera. The present invention can be similarly applied to recording and reproduction of video and the like. Further, the object to be searched is not limited to the characters, and may be other specific items such as food and goods.

  FIG. 1 schematically shows a configuration of a recording / reproducing apparatus according to an embodiment of the present invention. This recording / reproducing apparatus includes a recording device 11 as a recording means and a reproducing means, a remote controller (hereinafter simply referred to as a remote controller) 12, a display device 13 as a display means, and a face database 14 as a storage means. ing.

  The recording device 11 is connected to the display device 13 and is connected to the face database 14 via the Internet line 15 and receives operation information from the remote controller 12. For example, the recording device 11 is included in a TV broadcast program and the program. Information on the characters appearing is recorded, and the recorded program is reproduced. Further, face collation processing as will be described later is performed. As a recording medium, a magnetic tape, a hard disk, a DVD, or the like is used.

The remote controller 12 is used by the user M to instruct search, specify a program, specify a character, specify a collation result image, and the like, and its operation information is transmitted to the recording device 11 by infrared communication or the like.
The display device 13 displays a program recorded or reproduced by the recording device 11, displays characters, displays a face matching result, and the like.

  The face database 14 stores (registers) face images of characters appearing in the program in advance, and an example thereof is shown in FIG. According to FIG. 2, a character name (group name) and a file name of a face image of the character are registered. Here, the facial image of the character may be registered by, for example, reproducing the recorded program and extracting and registering the facial image of the character from the image being reproduced. It is also possible to prepare and read and register the still image with a scanner or the like. Furthermore, the face images of characters may be acquired and registered using the Internet.

Next, the operation in such a configuration will be described.
First, a recording process of program and character information by the recording device 11 will be described with reference to a flowchart shown in FIG. The program name to be recorded designated by the user M and the names of characters appearing in the program are acquired from an electronic program table (not shown) and recorded in the form of a table as shown in FIG. 4, for example (step). S1). According to FIG. 4, a program name, a character name, and a recording date / time are registered in association with each other. Next, a television broadcast is received based on the registered recording date and time, and a program with the registered program name is recorded (recorded) (step S2). Here, the processing of steps S1 and S2 corresponds to the recording means in the present invention.
This completes the process of recording information on one program and characters, and repeats the above operation if a plurality of program names to be recorded are registered.

  Next, a first processing example for reproducing a recorded program will be described with reference to the flowchart shown in FIG. When the user M sends a search instruction from the remote controller 12 (step S11), the recording apparatus 11 that has received it determines whether or not the program is currently being reproduced (step S12). A list of characters of the program being reproduced is displayed (step S13). That is, the names of the characters of the program being reproduced are extracted from the information of the characters shown in FIG. 4 registered in advance, and the list is displayed on the display device 13.

If the result of determination in step S12 is that the program is not being played back, a list of recorded programs is displayed (step S14). That is, the program name is extracted from the character information shown in FIG. 4 registered in advance, and the list is displayed on the display device 13.
Here, the user M designates the program name desired to be reproduced by operating the remote controller 12 from the program list displayed on the display device 13 (step S15). Receiving it, the recording apparatus 11 displays a list of characters of the designated program (step S16). That is, the names of the characters of the designated program are extracted from the information of the characters shown in FIG. 4 registered in advance, and the list is displayed on the display device 13.

  Here, the user M designates a character name to be searched from the list of characters displayed on the display device 13 by operating the remote controller 12 (step S17). Receiving this, the recording apparatus 11 acquires a face image of the designated character name from the face database 14 via the Internet line 15 (step S18). That is, the face image of the designated character name is acquired from the data of the face database 14 shown in FIG. 2 registered in advance. Here, the processing of steps S17 and S18 corresponds to the face image acquisition means (specific object image acquisition means) in the present invention.

  Next, the recording device 11 collates the face image acquired from the face database 14 with the face image of the person appearing in the program (step S19), and the collation match is obtained first from the time when the collation is started. The reproduction of the program is started from the point of time (step S20). Here, the processing in step S19 corresponds to the collating means in the present invention, and the processing in step S20 corresponds to the reproducing means in the present invention.

  Next, a second processing example for reproducing a recorded program will be described with reference to the flowchart shown in FIG. When the user M sends a search instruction from the remote controller 12 (step S211), the recording apparatus 11 that has received it determines whether or not the program is currently being reproduced (step S22). Then, a list of characters of the program being reproduced is displayed (step S23). That is, the names of the characters of the program being reproduced are extracted from the information of the characters shown in FIG. 4 registered in advance, and the list is displayed on the display device 13.

If the result of determination in step S22 is that the program is not being played back, a list of recorded programs is displayed (step S24). That is, the program name is extracted from the character information shown in FIG. 4 registered in advance, and the list is displayed on the display device 13.
Here, the user M designates a program name desired to be reproduced from the program list displayed on the display device 13 by operating the remote controller 12 (step S25). Receiving it, the recording apparatus 11 displays a list of characters of the designated program (step S26). That is, the names of the characters of the designated program are extracted from the information of the characters shown in FIG. 4 registered in advance, and the list is displayed on the display device 13.

  Here, the user M designates the name of the character to be searched from the list of characters displayed on the display device 13 by operating the remote controller 12 (step S27). Receiving this, the recording apparatus 11 acquires a face image of the designated character name from the face database 14 via the Internet line 15 (step S28). That is, the face image of the designated character name is acquired from the data of the face database 14 shown in FIG. 2 registered in advance. Here, the processing of steps S27 and S28 corresponds to the face image acquisition means (specific object image acquisition means) in the present invention.

  Next, the recording device 11 collates the face image acquired from the face database 14 with the face image of the person appearing in the program (step S29), and a list of images (scenes) at the time when the collation match is obtained. Is displayed on the display device 13 (step S30). Here, the process of step S29 corresponds to the collating means in the present invention, and the process of step S30 corresponds to the display means in the present invention.

  Here, the user M selects an image to be searched from the list displayed on the display device 13 by operating the remote controller 12 (step S31). Receiving it, the recording apparatus 11 starts reproduction of the program from the time point of the image selected in step S31 (step S32). Here, the processing of steps S31 and S32 corresponds to the reproducing means in the present invention.

  The above description is an example in which a program is reproduced from the time point of matching and matching. However, the program can be reproduced from a point of time (several seconds) after the point of matching and matching.

Next, the collation process in step S19 in FIG. 5 and step S29 in FIG. 6 will be described in detail.
First, a first example of collation processing will be described in detail with reference to the flowchart shown in FIG. In the first example of the collation process, the similarity between the face image to be searched and each face image extracted in the program is calculated, and at least one face image having a relatively high similarity is regarded as a collation matching image. It is what you regard.

  That is, first, face images of all persons appearing in the program to be reproduced are extracted (step S41). Next, the degree of similarity between all the extracted face images and the face image to be searched (face image acquired from the face database 14) is calculated (step S42). Next, the face image having the highest similarity among the calculated similarities is regarded as a matching image (step S43).

  Next, a second example of collation processing will be described in detail with reference to the flowchart shown in FIG. In the second example of the collation process, the similarity between the face image to be searched and the individual face images extracted in the program is calculated, and one face image having a predetermined similarity (for example, similarity 700) or more is calculated. If not, at least one face image having a relatively high degree of similarity (for example, the top five images having the highest degree of similarity) is regarded as a matching image.

  That is, first, face images of all persons appearing in the program to be reproduced are extracted (step S51). Next, the degree of similarity between all the extracted face images and the face image to be searched (the face image acquired from the face database 14) is calculated (step S52). Next, it is determined whether or not there is a face image having a predetermined similarity or higher (for example, a similarity of 700 or higher) among the calculated similarities (step S53). The image is regarded as a matching image (step S54), and if it does not exist, a predetermined number (for example, five) of face images from the top in the descending order of similarity are regarded as matching images (step S55).

  Next, a third example of collation processing will be described in detail with reference to the flowchart shown in FIG. In the third example of the collation processing, the similarity between the face image to be searched and the individual face images extracted in the program is calculated, and face images having a predetermined similarity (for example, similarity 700) or more are predetermined. When the number of images (for example, 20) or more exists, a certain number (for example, 5) of images in the descending order of similarity are regarded as matching images.

  That is, first, face images of all persons appearing in the program to be reproduced are extracted (step S61). Next, the degree of similarity between all the extracted face images and the face image to be searched (face image acquired from the face database 14) is calculated (step S62). Next, it is determined whether or not there are a predetermined number or more (for example, 20 or more) of face images having a predetermined similarity or more (for example, a similarity of 700 or more) among the calculated similarities (step S63). If present, a predetermined number (for example, five) of face images from the top in the descending order of similarity are regarded as matching images (step S64), and if not, face images having a predetermined similarity or higher are regarded as matching images. (Step S65).

1 is a schematic diagram schematically showing a configuration of a recording / reproducing apparatus according to an embodiment of the present invention. The figure which shows the example of the content of a face database. The flowchart explaining the recording process of the program and character information by a recording device. The figure which shows an example of the information of the character to record. The flowchart explaining the 1st process example which reproduces | regenerates the recorded program. The flowchart explaining the 2nd process example which reproduces | regenerates the recorded program. The flowchart explaining the 1st example of collation processing. The flowchart explaining the 2nd example of collation processing. The flowchart explaining the 3rd example of collation processing.

Explanation of symbols

  M ... user, 11 ... recording device (recording means and reproducing means), 12 ... remote controller, 13 ... display device (display means), 14 ... face database (storage means), 15 ... internet line.

Claims (11)

A recording means for recording information of a video and a specific object appearing in the video,
Storage means for storing an image of a specific object in advance;
Specific object image acquisition means for acquiring, from the storage means, an image of the specific object designated by the user based on the information of the specific object recorded by the recording means;
A collation unit that collates the image of the specific object acquired by the specific object image acquisition unit and the image of the specific object appearing in the video recorded by the recording unit, and outputs the verification result;
As a result of the collation by the collation means, if a match between both images is obtained, a reproduction means for reproducing the video recorded by the recording means from the time when the coincidence was first obtained after the collation was started,
A recording / reproducing apparatus comprising: A recording means for recording information of a video and a specific object appearing in the video,
Storage means for storing an image of a specific object in advance;
Specific object image acquisition means for acquiring, from the storage means, an image of the specific object designated by the user based on the information of the specific object recorded by the recording means;
A collation unit that collates the image of the specific object acquired by the specific object image acquisition unit and the image of the specific object appearing in the video recorded by the recording unit, and outputs the verification result;
As a result of the collation by the collation means, if a match between both images is obtained, a display means for displaying an image of the video at the time when the coincidence was obtained;
Playback means for playing back the video recorded by the recording means from the time of the image selected by the user from the image displayed by the display means;
A recording / reproducing apparatus comprising: Recording means for recording the received broadcast program and information on the characters appearing in the program;
Storage means for storing face images of characters in advance;
Face image acquisition means for acquiring from the storage means a facial image of the character designated by the user based on the information of the characters recorded by the recording means;
A collation unit that collates the face image acquired by the face image acquisition unit with the face image of the character appearing in the program recorded by the recording unit, and outputs the collation result;
As a result of the collation by the collation means, if a match between both face images is obtained, a reproduction means for reproducing the program recorded by the recording means from the time when the coincidence was first obtained after the collation was started,
A recording / reproducing apparatus comprising:   4. The recording / reproducing apparatus according to claim 3, wherein the reproducing means reproduces the program recorded by the recording means from a point in time when a match is obtained for a certain time. Recording means for recording the received broadcast program and information on the characters appearing in the program;
Storage means for storing face images of characters in advance;
Face image acquisition means for acquiring from the storage means a facial image of the character designated by the user based on the information of the characters recorded by the recording means;
A collation unit that collates the face image acquired by the face image acquisition unit with the face image of the character appearing in the program recorded by the recording unit, and outputs the collation result;
As a result of the collation by the collation means, when a match between both face images is obtained, a display means for displaying an image of the program at the time when the coincidence was obtained;
Playback means for playing back the program recorded by the recording means from the time of the image selected by the user from the image displayed by the display means;
A recording / reproducing apparatus comprising:   6. The recording / reproducing apparatus according to claim 5, wherein the reproducing means reproduces the program recorded by the recording means from a point in time when the user selects a certain time from an image point.   The collation means extracts the facial images of the characters from the program recorded by the recording means, and calculates the degree of similarity between the extracted individual facial images and the facial images acquired by the facial image acquisition means. 6. The recording / reproducing apparatus according to claim 3, wherein the face image having the highest similarity is regarded as a collation matching image.   The collation means extracts the facial images of the characters from the program recorded by the recording means, and calculates the degree of similarity between the extracted individual facial images and the facial images acquired by the facial image acquisition means. 6. The recording / reproducing apparatus according to claim 3, wherein when there are no face images having a predetermined similarity or higher, a predetermined number of face images are regarded as collation matching images in descending order of similarity. . The collation means extracts the facial images of the characters from the program recorded by the recording means, and calculates the degree of similarity between the extracted individual facial images and the facial images acquired by the facial image acquisition means. 6. When there are a predetermined number or more of face images having a predetermined similarity or higher, the predetermined number of face images from the top in the descending order of similarity are regarded as matching matching images. Recording / playback device.   From the storage means for recording the video and the information of the specific object appearing in the video, and storing the specific object image designated by the user based on the recorded specific object information in advance When the acquired image of the specific object and the image of the specific object appearing in the recorded video are collated and a match between the two images is obtained, the first match is found after the collation is started. A recording / reproducing method, wherein the recorded video is reproduced from the obtained time point.   From the storage means for recording the video and the information of the specific object appearing in the video, and storing the specific object image designated by the user based on the recorded specific object information in advance When the acquired image of the specific object and the image of the specific object appearing in the recorded video are collated and a match between both images is obtained, the image of the video at the time when the match is obtained And the recorded video is reproduced from the point of time of the image selected by the user from the displayed image.

Images