KR101927965B1 - System and method for producing video including advertisement pictures - Google Patents

Abstract

The present invention relates to a system and a method for producing advertisement videos, the method comprising: an original video is divided into a plurality of unit images; an event related feature is identified for each divided unit image and an event type corresponding to the event related feature is identified for each unit image; metadata of each unit image is generated and the advertisement image corresponding to the event type is inserted to a corresponding unit image using the corresponding metadata for the unit image including information of the event type in the corresponding metadata; and information for creating advertisement images corresponding to the event type is edited by using the corresponding metadata to generate an advertisement video.

Description

TECHNICAL FIELD [0001] The present invention relates to a system and a method for producing an advertisement video,

The present invention relates to a video advertisement production system and method, and more particularly, to a system and a method for producing an advertisement video by inserting an advertisement video into various types of videos such as UCC (User Created Contents), movies, dramas, music videos, .

Videos such as UCC (User Created Contents), movies, dramas, music videos, and animations are provided through a broadcasting medium such as a TV, or are presented to a viewer through a theater or through the Internet.

When a video is provided for the purpose of profit, most of the users receive the advertisement image from the provider of the advertisement image before the start of the video, after the end of the video, or during the watching of the video.

However, most of the commercials are irrelevant to the content of the video, and even if they are relevant, most of the videos related to the whole topic of the video or the actor. For this reason, the conventional advertisement image is distant from the impressions given by the moving images or the event conditions of the specific scenes, and the synergy effect with the moving images can not be provided.

Conventionally, a method of providing an advertisement image while viewing a moving image is a method of exposing a designated advertisement image irrelevant to a scene being viewed immediately before pausing a moving image. Therefore, when a viewer interferes with the immersion of a moving image and gives an uncomfortable feeling Are frequent.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a system and a method for producing an advertisement moving image which has a high advertising effect by exposing an advertisement image having a synergy effect with a moving image during watching a moving image.

It is another object of the present invention to provide a system and method for producing an advertisement moving image in which an advertisement image corresponding to an event of each scene of a moving image is automatically matched and selected and an advertisement moving image is generated.

Another object of the present invention is to provide a system and method for creating an advertisement moving image in which an advertisement image exposed during a moving image has similarity with a previous scene, thereby preventing a viewer from having a sense of heterogeneity with respect to the advertisement image.

Embodiments according to the present invention can be used to accomplish other tasks not specifically mentioned other than the above-described tasks.

According to an embodiment of the present invention, there is provided an advertisement moving image production system comprising: a unit image feature extraction unit for dividing an original moving image into a plurality of unit images and recognizing an event association feature for each divided unit image; An event unit for recognizing an event type corresponding to an event relevance feature for each unit image, and metadata for each unit image using a signal received from the event unit, A meta information section including information for judging whether or not to edit the content and information used at the time of inserting or editing, a advertisement image information storage section for storing information for creating an advertisement image or an advertisement image corresponding to an event type, In the metadata, a unit image including information on an event type The advertisement image corresponding to the type is inserted into the unit image using the corresponding metadata or the information for creating the advertisement image corresponding to the event type is edited using the corresponding metadata to generate the advertisement video And an advertisement moving picture generating unit.

The event association feature is a feature word that is obtained by analyzing video, audio, and subtitles of a unit image, and the feature word is a word stored in the storage unit.

The unit image is an image segmented based on the set time or divided into a plurality of identical image frames.

The unit image feature recognition unit recognizes the background color in the image of each image frame forming the unit image, and identifies the small items to further grasp the background relevance feature.

The background relevance feature is included in the meta data, and the advertisement video generation unit has the advertisement image that is inserted or edited according to the background relevance feature have a video background suited to the background association.

The event unit identifies one event type corresponding to the feature word for each unit image, and selects an event type corresponding to the feature word having the highest exposure frequency among the event types if the event type is two or more for one unit image .

The event unit classifies an event type corresponding to each unit image according to an event type, and determines a priority according to a set priority condition if the event type is two or more.

And the advertisement video generation unit may insert the advertisement video corresponding to the event type into the unit video having the highest priority among the unit videos having the same event type, The unit image is edited and made into an advertisement image.

The priority condition includes a first condition in which an event related feature word has a high exposure frequency, a second condition in which a unit image having the same event exists immediately before and after the unit image, a third condition in which a scene corresponds to a condition requested by the advertiser, The fourth condition, which is a scene in which the purchasing power of the product is high, and the fifth condition, in which the content in the moving picture is a climax. Alternatively, the priority condition may be a first condition in which an event related feature word has a high exposure frequency, a second condition in which a unit image having the same event exists immediately before and after the unit image, a third condition A fourth condition which is a scene in which the purchasing power of the advertisement product is high, and a fifth condition in which the content in the moving picture is a climax are satisfactory.

The unit image feature recognition unit recognizes the position of each character and each character in the image of each image frame constituting the unit image and recognizes the character having the highest exposure frequency among the characters or the character located at the center of the screen as a main character, The location of the character and the position of the face, and the position of the main character and the position of the face for each unit image are included in the metadata.

The metadata includes conversion information for converting an advertisement image or a unit image to be inserted into an advertisement image, and the conversion information includes information on the position of the main character and the position of the face for each unit image, It is information for changing into a face.

According to another aspect of the present invention, there is provided a method for producing an advertisement moving picture, including dividing an original moving picture into a plurality of unit images, The method comprising the steps of: identifying an event type corresponding to an event relevance feature for each image; generating metadata for each unit image, the metadata including information for determining whether to insert or edit the advertisement image in the unit image, A step of inserting the advertisement image corresponding to the event type into the corresponding unit image by using the corresponding metadata, Information for creating an advertisement image corresponding to an event type is referred to as " And a step of generating an advertisement video by editing the image corresponding unit.

Here, the event-related characteristic is a characteristic word that is obtained by analyzing video, audio, and subtitles of a unit image. The characteristic word is a previously stored word, and the unit image may be a divided image based on the set time, It is an image divided into the same image frame.

In the step of recognizing the event-related characteristic, the background color is recognized in each image frame image forming the unit image, and the small objects are identified to grasp the background-related characteristic.

The background relevance feature is included in the metadata and the step of generating the advertisement moving image has an image background that is inserted or edited according to the background relevance feature matches the background association.

The step of determining the type of the event identifies one event type corresponding to the feature word for each unit image. If the event type is two or more, the event type of the feature word having the largest exposure frequency among the event types is selected.

The step of recognizing the event type classifies the event type corresponding to each unit image according to the event type, and if the event type is two or more, the priority is determined according to the set priority condition.

Wherein the step of generating the advertisement moving picture comprises the step of displaying the advertisement image corresponding to the event type in the unit video having the highest priority among the unit videos having the same event type Or the unit image is edited to produce an advertisement image.

The priority condition includes a first condition in which an event related feature word has a high exposure frequency, a second condition in which a unit image having the same event exists immediately before and after the unit image, a third condition in which a scene corresponds to a condition requested by the advertiser, The fourth condition, which is a scene in which the purchasing power of the product is high, and the fifth condition, in which the content in the moving picture is a climax. Alternatively, the priority condition may be a first condition in which an event related feature word has a high exposure frequency, a second condition in which a unit image having the same event exists immediately before and after the unit image, a third condition A fourth condition which is a scene in which the purchasing power of the advertisement product is high, and a fifth condition in which the content in the moving picture is a climax are satisfactory.

The step of recognizing the event relatedness feature may be performed by grasping the position of each character and each character in the image of each image frame constituting the unit image and recognizing the character located at the center of the screen or the screen having the highest frequency of exposure as a main character , The position of the main character and the position of the face, and the position of the main character and the position of the face for each unit image are included in the metadata.

The metadata includes conversion information for converting an advertisement image or a unit image to be inserted into an advertisement image, and the conversion information includes information on the position of the main character and the position of the face for each unit image, It is information for changing into a face.

According to the embodiment of the present invention, the advertisement image having the synergy effect with the moving picture being displayed is exposed during the watching of the moving picture, thereby having a high advertising effect.

Also, according to the embodiment of the present invention, the viewer can maintain the immersion level of the moving image by preventing the viewer from having a sense of heterogeneity with respect to the advertisement image exposed during watching the moving image.

1 is a conceptual diagram of an advertising moving image production system and method according to a first embodiment of the present invention.
2 is a conceptual diagram of an advertising moving image production system and method according to a second embodiment of the present invention.
3 is a block diagram of an advertisement moving image production system according to an embodiment of the present invention.
4 is a detailed configuration diagram of a unit image feature determination unit according to an embodiment of the present invention.
5 is a detailed configuration diagram of an image feature determination unit according to an embodiment of the present invention.
6 is a detailed configuration diagram of an event unit and a meta information unit according to an embodiment of the present invention.
7 is a detailed configuration diagram of an advertisement moving image generating unit according to the first embodiment of the present invention.
8 is a detailed configuration diagram of an advertisement moving image generating unit according to the second embodiment of the present invention.
FIG. 9 is a flowchart illustrating an advertisement moving picture production method according to the first embodiment of the present invention.
FIG. 10 is a flowchart illustrating an advertisement moving picture production method according to a second embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings, which will be readily apparent to those skilled in the art to which the present invention pertains. The present invention may be embodied in many different forms and is not limited to the embodiments described herein. In order to clearly illustrate the present invention, parts not related to the description are omitted, and the same reference numerals are used for the same or similar components throughout the specification. In the case of publicly known technologies, a detailed description thereof will be omitted.

In this specification, when a part is referred to as "including " an element, it is to be understood that it may include other elements as well, without departing from the other elements unless specifically stated otherwise. Also, the terms "part," " module, "and the like, which are described in the specification, refer to a unit for processing at least one function or operation, and may be implemented by hardware or software or a combination of hardware and software.

Hereinafter, a system and method for producing an advertisement moving image according to an embodiment of the present invention will be described with reference to the accompanying drawings.

First, the concept of the present invention will be described with reference to Figs. 1 and 2. Fig.

1 is a conceptual diagram of an advertising moving image production system and method according to a first embodiment of the present invention. Referring to FIG. 1, an advertisement moving image production system and method according to an embodiment of the present invention includes: (a) inserting an advertisement image into at least one part of an original image as shown in (b) To produce an advertisement video.

As a result of inserting the advertisement image, the length (viewing time) of the produced advertisement video becomes longer than that of the original image. That is, the length (viewing time) of the advertisement video is the length (viewing time) obtained by adding the length (viewing time) of the advertisement image to the length of the original image (viewing time). The advertisement video described above and below means a video including an advertisement video on the original video.

However, if the advertisement image is randomly inserted in the middle of the original image, the viewer immediately recognizes that the advertisement image is the advertisement image, and the advertisement effect is not related to the content being watched. Therefore, the present invention generates an advertisement video by inserting an advertisement image correlated to an A scene or a B scene of an original image to be inserted, or generates an advertisement video image-processed to have relevance to an A scene or a B scene of the original image.

Here, the association between the A scene or the B scene of the original image to be inserted and the advertisement image is classified into the event association and the background association.

Event associations are related to the type of advertising image. That is, the event association is related to one of advertisement (eg, car, necklace, computer, personal property, etc.) appearing in scene A or scene B, (Such as a birthday, a birthday of a parent, a baby's birthday party, a trip, a beachfront, a mountaineering, a traffic accident, or the like), or related to a main character . Of course, the event association includes not only the above description but also the word related to the word coming from the scene.

The background relevance is a relationship between the background of the advertisement image and the background of the A scene or the B scene, in order that the user naturally recognizes the advertisement image during the viewing of the image, that is, the advertisement image is naturally synchronized with the inserted scene. For example, a scene A and a scene B have the background of a beach, and the advertisement image has a background of mountains or a background of a sunny day, and the advertisement image has a low background or an indoor background. The viewer can easily understand that the viewer is the advertisement image, and the immersion of the viewer of the video is deteriorated. Therefore, the background association includes making the scene background of the original image similar to the scene background of the advertisement image.

2 is a conceptual diagram of an advertising moving image production system and method according to a second embodiment of the present invention. Referring to FIG. 2, an advertisement moving image production system and method according to an embodiment of the present invention may be configured such that, as shown in (b) of an original image shown in (a) Produce a movie. Accordingly, unlike the first concept, the concept of the second embodiment of the present invention is that the length (viewing time) of the advertisement moving picture is the same as the length of the original image (viewing time) although the advertisement image is included. That is, even if the advertisement image is included, the length of the original image does not change.

Editing a part of the image into an advertisement image means that the set advertisement product is included in the image frame and the main character or the face of the main character in the image frame, the specific object or the background other than the main character are changed. For example, in a video segment where a scene of playing a cola in a car running behind a nice background is played, the car is changed to "A's car", the coke is changed to "B's bottled water" The advertisement signboard is displayed, and the main character (character) is replaced with the character of the third party to be reproduced.

Here, the character of the third party may be an advertiser, a person who made a UCC, or a viewer who is currently watching the corresponding video. In the case of an advertiser, the photograph of the face of the advertiser or other photographs are provided in advance and utilized. If the person who created the UCC is provided with the UCC, Before the viewer watches the video, he or she can identify the face photograph or other photo of the registered viewer by providing face photographs or other photographs or by authenticating the user through login.

In the second concept according to the embodiment of the present invention, if the advertisement image is randomly inserted in the middle of the original image, the viewer immediately recognizes that the advertisement image is the advertisement image, . Therefore, the present invention inserts an advertisement image having event relevance and background relevance in an A scene or B scene of an original image into which an advertisement image is inserted.

Background relationships in the first and second concepts described above may be excluded in some cases.

Hereinafter, an advertisement moving image production system according to an embodiment of the present invention for achieving the first or second concept of the present invention will be described with reference to FIG.

3 is a block diagram of an advertisement moving image production system according to an embodiment of the present invention. 3, an advertisement moving image production system according to an embodiment of the present invention includes an original moving image storage unit 110, a target image extracting unit 120, a unit image feature estimating unit 130, an event unit 140, An advertisement information storing unit 160, an advertisement moving image generating unit 170, and an advertisement moving image storing unit 180. The information display unit 150 includes an advertisement image information storing unit 160,

The original moving picture storage unit 110 stores original moving pictures such as UCC, movie, and drama. The original video means a video in which the advertisement video is not inserted.

The target image extracting unit 120 selects one original moving image corresponding to the received extraction command from the original moving image storing unit 110. [ Here, the extraction command may be an instruction automatically generated by an event stored in the original moving picture storage unit 110, an instruction automatically generated by an event generated when a viewer requests viewing of the original moving picture, An order for production of an advertisement video by the producer, and the like.

The unit image feature determination unit 130 divides one original moving image extracted by the target image extraction unit 120 into a plurality of unit images and determines a feature (hereinafter referred to as " Event association feature "). If the background relevance is required, the unit image feature determination unit 130 searches for a feature (hereinafter, referred to as "background relevance feature") related to the background association for each unit image. Hereinafter, the event-related characteristics of each unit image and the background-related features of each unit image are collectively referred to as "features per unit image".

The event association feature is a word associated with a props, a word related to a situation, or a word related to a main character. The background relevance feature is information about a place (indoor, outdoor, swimming pool, park, pub), background (night or day, etc.).

A unit image is divided into images based on the same time such as 1 second, 2 seconds, and 5 seconds, and is composed of a plurality of image frames. Alternatively, the unit image may be an image divided into the same number of image frames, such as 10, 20, 50, and so on.

The event unit 140 analyzes the feature of each unit image detected by the unit image feature detection unit 130 to determine whether or not an event related to the feature of each unit image is generated, , The same event is classified into each unit image, and when a plurality of the same events are classified, a priority unit image is determined, and the determined priority is flagged for each unit image.

The meta information unit 150 generates metadata for each unit image using the information received from the event unit 140 and the unit image feature determination unit 130. [ The meta data includes information for determining whether to insert or edit the advertisement image in the unit video, and information used in the insertion or editing. For example, the meta data is information for determining whether to insert or edit the advertisement image in the unit video, and includes the event occurrence information and the event priority. The metadata is information used at the time of insertion or editing, and includes an event type and conversion information. The conversion information will be described in detail in the following description with reference to FIG.

The advertisement image information storage unit 160 stores advertisement images requested by advertisers, words related to each advertisement image, and information about the screen background in the case of applying to the first concept of the present invention, Respectively. Alternatively, when applied to the second concept of the present invention, the advertisement image information storage unit 160 stores the template information of each object for the corresponding advertisement image corresponding to the event type.

The advertisement moving image generating unit 170 grasps the related advertisement image based on the meta data of the corresponding unit image received from the meta information unit 150 for each unit image and inserts or edits the advertisement image when there is an associated advertisement image And generates an advertisement video. The advertisement video storage unit 180 stores the advertisement video generated by the advertisement video generation unit 170.

Hereinafter, each configuration of the advertisement moving image production system according to the embodiment of the present invention will be described in more detail with reference to FIG. 4 to FIG.

First, the unit image characteristic determination unit 130 will be described in detail with reference to FIG. 4 is a detailed configuration diagram of a unit image feature determination unit according to an embodiment of the present invention.

4, the unit image feature determination unit 130 according to the embodiment of the present invention includes a division unit 131, an object separation unit 132, an image feature detection unit 133, a voice feature detection unit 134, And a caption feature determination unit 135. [0050]

The division unit 131 divides one original moving image extracted by the target image extraction unit 120 into a plurality of unit images and provides the divided unit images to the object separation unit 132. [ The object separating unit 132 separates each image frame of the received unit image into an object such as an image (video), audio, and a caption (text), provides the separated image to the image feature determining unit 133, One audio is provided to the voice feature detection unit 134, and the separated caption is provided to the caption feature detection unit 135.

The image feature determination unit 133 receives an image for each image frame constituting a unit image and analyzes the image to identify an event association feature, that is, a feature word related to the event. Here, the image feature determination unit 133 may analyze the image of each image frame forming the unit image to further grasp the background relevance feature.

The image characteristic determination unit 133 will be described in more detail with reference to FIG. 5 is a detailed configuration diagram of an image feature determination unit according to an embodiment of the present invention.

5, the image feature determination unit 133 includes a screen background determination unit 133a, a main character recognition unit 133b, a screen object recognition unit 133c, and a feature recognition unit 133d.

The screen background grasping unit 133a grasps the color of the background image, that is, the background color, and provides the information on the background color to the feature grasping unit 133d. The main character grasping unit 133b grasps the position of each character and each character in the image of each video frame constituting the unit video, and recognizes the character having the greatest exposure frequency among the characters and / or the character positioned at the center of the screen Character, recognizes the face of the main character, and provides the position of the main character and the position of the face to the characterizing unit 133d.

The screen object grasping unit 133c grasps a small object (object) in each image frame image forming the unit image and identifies the type of small object (e.g., a chair, a sofa, a table, a bed, a hat, a cake, And supplies the kind of the small article, that is, the word of the small article, to the characterizing section 133d. On the other hand, when the unit image is edited into the advertisement image, the screen object grasping unit 133c grasps the position of the object and the small object in the image of each image frame constituting the unit image, , A table, a bed, a hat, a cake, an automobile, etc.), and provides the characterizing unit 133d with the type of the identified small object, that is, the word and position of the small object.

The feature determination unit 133d sets the feature word in the feature word storage unit 136 among the unit image-based words received from the screen object recognition unit 133c as a feature word and sets the feature word as information on the event association feature And provides it to the event unit 140. In this case, one or more feature words may be used per unit image. The feature detection unit 133d detects the background correlation feature (e.g., indoor, office, park, road, etc.) of the housing through the word of the small article received from the screen object recognition unit 133d, (150). Here, when the unit image is edited into the advertisement image, the feature identification unit 133 provides the meta information unit 150 with the background relevance feature and the position of the word and the props of the props per unit image.

The feature determination unit 133d also provides the meta information unit 150 with information on the main character (i.e., the position of the main character and the position of the face) from the main character recognition unit 133b.

The main character grasping unit 133b and the screen object grasping unit 133c in the image feature grasping unit 133 can be omitted by the maker.

The voice feature recognition unit 134 receives the audio or voice for each video frame constituting the unit video, recognizes the word through the voice recognition function, and recognizes the word stored in the feature word storage unit 136 Identify by feature word.

The caption feature detection unit 135 receives the caption of the text of each video frame constituting the unit video and extracts a word that is a noun in the caption and extracts a word stored in the feature word storage unit 136 among the extracted words, It is understood as a word. In the above, there may be more than one feature word for each unit image.

The feature word storage unit 136 stores a plurality of noun words. It is preferable that the words stored in the feature word storing unit 136 are words related to the advertisements ordered from the advertiser and that they are stored or deleted in the feature word storing unit 136 in synchronization with the receipt of a new advertisement or termination of an advertisement contract Do.

Hereinafter, the event unit 140 and the meta information unit 150 will be described in more detail with reference to FIG. 6 is a detailed configuration diagram of an event unit and a meta information unit according to an embodiment of the present invention.

6, the event unit 140 includes an event determination unit 141, a priority determination unit 142, and an image tag processing unit 143. [

The event recognition unit 141 uses the feature words of each unit image received from the video feature recognition unit 133, the voice feature recognition unit 134 and the caption feature recognition unit 135 of the unit image feature recognition unit 130 And determines the event occurrence and the kind of the event. For example, when the event determination unit 141 determines that an event has occurred, the type of the event including the feature word or the feature word is determined. If the type of the detected event is two or more, Selects the type of the event of the feature word having the largest number of the feature words, and matches the type of the selected event with the identification information of the corresponding unit image. Here, the type of the event may be birthday, joining, sixtieth birthday, wedding, travel, etc., and the characteristic word is a cake, a bouquet, a birthday present,

It is determined that an event has not occurred with respect to a unit image having no feature word, and it is determined that an event has occurred with respect to a unit image having the feature word. In another example, the event detection unit 141 may determine whether or not the feature frequency of the feature word is equal to or less than the preset frequency even if the feature word is present, or the feature word of the feature recognition unit 133 and the feature of the voice feature recognition unit 134 Determination unit) is not a word of the same event type, it is determined that no event has occurred.

The priority determining unit 142 determines a priority order for a plurality of unit videos having the same event type. To this end, the priority determining unit 142 receives the event type for each unit video having the event, classifies the unit video according to the same event type, and then determines the priority according to the set priority condition of the unit video for each event type. Here, the priority condition includes a first condition in which an event related characteristic word has a high frequency of exposure, a second condition in which a unit image having the same event exists immediately before and after the unit image, a third condition in which a scene corresponds to a condition requested by the advertiser, A fourth condition which is a scene in which the purchasing power of the product is high, and a fifth condition that the content in the moving picture is a climax.

Of course, only one of the first to fifth conditions may be used, or at least two of the first to fifth conditions may be used. In addition, the weights may be different from each other for the first to fifth conditions, so that the priority can be made higher when the condition of high weight is satisfied.

The image tag processing unit 143 displays the tag value corresponding to the priority determined by the priority determining unit 142 on the corresponding unit image.

Next, the meta information unit 150 includes a conversion information acquisition unit 151 and a meta data generation unit 152. [ The conversion information acquiring unit 151 acquires the background relevance characteristic of each unit image, the information about the main character per unit image, received from the image feature acquiring unit 133, and the priority of each unit image received from the priority determining unit 142 (Or an upper set number) having the highest priority for each event type and generates conversion information for converting the advertisement image corresponding to the event type of the first unitary image. Herein, the conversion information is a background relevance feature for allowing the advertisement image to be a background, a background, and a situation similar to the previous image and the subsequent image at the position where the advertisement image is located in the original video. In addition, It is the information to change to the face of major character or main character of the image, to change to the face of the advertiser, to the face of the member watching the original video, or to the face of the producer who made the UCC in case of UCC. In addition, when the unit image is edited with the advertisement image, the conversion information further includes position information of the props and accessories related to the advertisement image in the unit image.

For example, the conversion information includes background color information corresponding to the unit video, information on the surrounding environment of the unit video (e.g., office, living room, road, beachfront, etc.), and when the character or character face is converted, The position information of the character corresponding to the image or the position information of the character face.

The metadata generating unit 152 receives the conversion information per unit image received from the conversion information acquiring unit 151 and the priority information received from the priority determining unit 142 and generates and outputs metadata for each unit image . Here, it is preferable that the unit image in which the metadata is generated is a unit image in which the advertisement image is to be inserted, and the unit image having the highest priority in each event type or the unit image in which the priority is the upper set number in each event type.

Hereinafter, the advertisement moving image generating unit 170 will be described in more detail with reference to FIG. 7 is a detailed configuration diagram of an advertisement moving image generating unit according to the first embodiment of the present invention.

Referring to FIG. 7, the advertisement moving image generating unit 170 according to the first embodiment of the present invention includes a matching advertisement extracting unit 171, an advertisement image processing unit 172, and a advertisement image inserting unit 173.

The matching advertisement extracting unit 171 sequentially receives the meta data for each unit image from the meta information unit 150 and analyzes the contents of the meta data to identify the type of event of the unit image and obtain an advertisement image corresponding to the detected event type From the advertisement image information storage unit 160.

The advertisement image processing unit 172 receives the advertisement image corresponding to the unit image from the matching advertisement extracting unit 171, receives the metadata of the unit image, and displays the received advertisement image according to the conversion information included in the metadata . At this time, if the main character of the advertisement image or the face of the main character is to be converted, the advertisement image processor 172 stores information on the character to be converted or the face of the character, or the character or character to be converted into the received advertisement image It contains information about the face.

The advertisement image inserting unit 173 receives the advertisement image processed by the advertisement image processing unit 172, receives the metadata of the unit image and the original moving image, Alternatively, an advertisement image is inserted into the unit image as shown in FIG. The created advertisement video is stored in the advertisement video storage unit 180.

8 is a detailed configuration diagram of an advertisement moving image generating unit according to the second embodiment of the present invention. Referring to FIG. 8, the advertisement moving image generating unit 170 according to the second embodiment of the present invention includes a matching advertisement extracting unit 171a and an original image editing unit 173a.

The matching advertisement extracting unit 171a sequentially receives meta data for each unit image from the meta information unit 150 and analyzes the contents of the meta data to generate a template (an object image, a character, Or a character face).

The original image editing unit 173a receives the template of the advertisement image corresponding to the unit image from the matching advertisement extracting unit 171, receives the metadata of the unit image and the original moving image, The unit image of the advertisement image is included in the unit image of the advertisement image, or the object of the unit image is replaced with the object of the advertisement image so that it becomes the same as the object of the advertisement image.

Hereinafter, a method for producing an advertisement moving picture according to an embodiment of the present invention will be described with reference to FIGS. 9 and 10. FIG.

FIG. 9 is a flowchart illustrating an advertisement moving picture production method according to the first embodiment of the present invention. As shown in FIG. 9, an advertisement moving picture production method according to the first embodiment of the present invention first extracts one original moving picture (S901), and divides the extracted original moving picture into unit images of a predetermined time length or number of frames (S902).

Then, each image frame constituting one unit image is separated into video, audio (audio) and subtitles (S903), and then the background relevance characteristic and word characteristic for the video, word characteristic for audio and word characteristic for caption (S904).

Then, it is determined whether there is an event included in the corresponding unit image (S905). If there is no event, the unit image of the next sequence is executed in steps S903 to S905. If there is any kind of event, it is identified and the kind of the detected event is matched with the unit image and stored (S907). Here, the type of the event is limited to those related to the advertisement image to be advertised.

Steps S903 to S907 are performed for all unit images (S908).

In steps S903 to S908, all unit images are classified into one of a first unit image having no event and a second unit image having an event, and the second unit images are classified again according to the same event type (S910).

The second unit images of the same event type are prioritized based on the background relevance feature and feature word, or priorities are determined in addition to the additional conditions (S911).

The reason for determining the priority is that there are many unit images with events for the original video. If the advertisement image is inserted for every unit image in which all the events are generated, it will cause a lot of irritation to viewers. Therefore, it is necessary to expose one, two, or a set number of advertisement images according to the same event type. For this purpose, the advertisement image is exposed corresponding to a unit image having a high priority.

If the priority of the second unit images of the same event type is determined and the advertisement image is exposed corresponding to the third unit image having the highest priority, for example, Meta data is generated for the unit image (S912).

In step S913, the metadata included in the meta data is analyzed for each of the third unit images, and an advertisement image associated with the corresponding third unit image is selected (extracted) in step S914. , The extracted advertisement image is inserted in front of or behind the corresponding third unit image to create an advertisement video (S915), and the advertisement video is stored (S916).

Here, for the more advanced advertisement moving image, the advertisement image may be further processed to have a screen background similar to the third unit image.

Hereinafter, a method of producing an advertisement moving picture according to a second embodiment of the present invention will be described with reference to FIG. FIG. 10 is a flowchart illustrating an advertisement moving picture production method according to a second embodiment of the present invention.

Referring to FIG. 10, steps S1001 to S1013 of the method for producing an advertisement moving image according to the second embodiment of the present invention are the same as those of the method for producing an advertisement moving image according to the first embodiment of the present invention (see S901 to S913) .

However, in the method of producing an advertisement moving picture according to the second embodiment of the present invention, an advertisement image template is extracted not from the advertisement image through the metadata per unit image (S1014), and the corresponding unit image (S1015), and storing the created advertisement video (S1016) is different from the first embodiment in that the unit video is edited so that the advertisement video template is included in the advertisement video template.

In the second embodiment of the present invention, since editing is performed on the original moving picture, there is an advantage in that it is unnecessary to grasp the background or surrounding environment information compared to the first embodiment.

While the present invention has been particularly shown and described with reference to exemplary embodiments, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, It belongs to the scope.

110: Original moving picture storage unit 120:
130: unit image feature recognition unit 140:
150: meta information unit 160: advertisement image information storage unit
170: advertisement video generation unit 180: advertisement video storage unit
131: partitioning unit 132: object separating unit
133: image feature detection unit 134: voice feature detection unit
135: caption feature recognition unit 136: feature word storage unit
133a: screen background grasp part 133b: main character grasping part
133c: Screen object grasping unit 133d:
141: Event detection unit 142: Priority determination unit
143: video tag processing unit 151: conversion information holding unit
152: Meta data generation unit 171, 171a: Matching advertisement extraction unit
172: an advertisement image processing unit 173:
173a: Original image editor

Claims (24)

A unit image feature extracting unit for dividing one original video into a plurality of unit videos and recognizing an event correlation feature for each divided unit video,
An event unit for identifying an event type corresponding to an event relevance feature for each unit image and classifying the event type according to an event type and for determining a priority order of a plurality of unit images having the same event type according to a set priority condition,
And generates metadata for each unit image using the signal received from the event unit. The metadata may include information for determining whether to insert or edit the advertisement image in the unit image, A meta information section including information,
An advertisement image information storage unit storing information for creating an advertisement image or an advertisement image corresponding to an event type, and
It is possible to insert an advertisement image corresponding to the event type into a unit image using the metadata and to create an advertisement image corresponding to the event type, And an advertisement moving picture generating unit for generating an advertisement moving picture by editing the corresponding unit video using the corresponding metadata,
Information on priority of each unit image is included in the metadata,
The priority condition includes a second condition in which a unit image having the same event exists immediately before and after the unit image, and a fourth condition in which a purchasing power of the advertisement product is high,
Wherein the event unit sets the priority of the unit video image to a higher priority if the number of conditions satisfying the priority condition is greater,
Wherein the advertisement video generation unit inserts an advertisement video corresponding to an event type into a unit video having the highest priority among the unit videos having the same event type or edits the unit video to create an advertisement video. The method of claim 1,
Wherein the event association feature is a feature word that is obtained by analyzing video, audio, and subtitles of a unit image, and the feature word is a word stored in a storage unit. The method of claim 1,
Wherein the unit video is a video segmented based on the set time or divided into a plurality of identical video frames. The method of claim 1,
Wherein the unit image feature determination unit identifies a background color in each image frame image forming the unit image and identifies the small items to further grasp the background relevance feature. 5. The method of claim 4,
Wherein the background relevance feature is included in the metadata,
Wherein the advertisement moving image generating unit has an advertisement image inserted or edited according to the background relevance feature having a video background suited to the background association. 3. The method of claim 2,
The event unit identifies one event type corresponding to a feature word for each unit image, and selects an event type corresponding to a feature word having the highest exposure frequency among the event types when the event type is two or more for one unit image Advertising video production system. delete delete The method of claim 1,
Wherein the priority condition further includes a first condition having a high exposure frequency of an event related characteristic word, a third condition being a scene corresponding to a condition requested by the advertiser, and a fifth condition in which the content in the image is a climax. delete The method of claim 1,
The unit image feature recognition unit recognizes the position of each character and each character in the image of each image frame constituting the unit image and recognizes the character having the highest exposure frequency among the characters or the character located at the center of the screen as a main character, The position of the character and the position of the face,
Wherein the position of the main character and the position of the face of each unit video are included in the metadata. 12. The method of claim 11,
The metadata includes conversion information for converting an advertisement image or a unit image to be inserted into an advertisement image, and the conversion information includes information on the position of the main character and the position of the face for each unit image, An advertisement video production system which is information for changing into a face. Dividing one original moving image into a plurality of unit images,
Determining an event association characteristic for each divided unit image,
Determining a priority order according to a set priority condition for a plurality of unit images having the same kind of event by classifying the event types corresponding to the event association characteristic for each unit video into the respective event types,
Wherein the meta data includes information for determining whether to insert or edit the advertisement image in the corresponding unitary image and information used in the insertion or editing,
It is possible to insert an advertisement image corresponding to the event type into a unit image using the metadata and to create an advertisement image corresponding to the event type, And editing the unit video using the metadata to generate an advertisement moving picture,
Information on priority of each unit image is included in the metadata,
The priority condition includes a second condition in which a unit image having the same event exists immediately before and after the unit image, and a fourth condition in which a purchasing power of the advertisement product is high,
Wherein when the number of conditions satisfying the priority condition is high, the priority of the unit video is set to a high priority,
Wherein the step of generating the advertisement moving image includes inserting the advertisement image corresponding to the event type into the unit image having the highest priority among the unit images having the same event type or editing the corresponding unit image into the advertisement image. The method of claim 13,
Wherein the event association feature is a feature word that is obtained by analyzing video, audio, and subtitles of a unit image, and the feature word is a previously stored word. The method of claim 13,
Wherein the unitary image is a divided image based on the set time or divided into a plurality of identical image frames. The method of claim 13,
Wherein the step of recognizing the event association feature grasps the background color in each video frame image constituting the unit video and identifies the small items to further grasp the background relevance characteristic. 17. The method of claim 16,
Wherein the background relevance feature is included in the metadata,
Wherein the step of generating the advertisement moving image comprises the step of causing the advertisement image inserted or edited according to the background relevance feature to have a video background suited to the background association. The method of claim 14,
The step of recognizing the type of the event may identify one kind of event corresponding to the feature word for each unit image and if the type of the event is two or more, Production method. delete delete The method of claim 13,
Wherein the priority condition further includes a first condition having a high exposure frequency of an event related characteristic word, a third condition being a scene corresponding to a condition requested by the advertiser, and a fifth condition that the content in the moving picture is a climax. delete The method of claim 13,
The step of recognizing the event relatedness feature may be performed by grasping the position of each character and each character in the image of each image frame constituting the unit image and recognizing the character located at the center of the screen or the screen having the highest frequency of exposure as a main character , The position of the main character and the position of the face,
Wherein the location of the main character and the position of the face for each unit video are included in the metadata. 24. The method of claim 23,
The metadata includes conversion information for converting an advertisement image or a unit image to be inserted into an advertisement image, and the conversion information includes information on the position of the main character and the position of the face for each unit image, How to make an advertisement video that is information for changing to a face.

Images