KR102538058B1 - Announcing advertisement banner provision system for website - Google Patents

Abstract

An announcing advertisement banner provision system for a website according to the present invention comprises: a banner creation module which includes an image input unit receiving an image including a performer's face, a voice input unit receiving informational voice including guidance information, a background removal unit removing a background from the image, and a banner generator creating an announcing banner by dubbing the informational voice to the image from which the background is removed; a location setting module which sets an output location of the announcing banner on a website; and a banner exposure module which displays and replays the announcing banner at the output location. According to the announcing advertisement banner provision system for a website of the present invention, a performer such as an announcer provides advertisements by displaying an announcing banner on the website that appears to provide advertisements in the form of voice, thereby increasing effectiveness of advertising for advertisement viewers who may be users of the website where a relevant announcing banner is inserted.

Description

Announcing advertisement banner provision system for website}

The present invention relates to a system for providing an announcement advertisement banner for a website, and more specifically, an announcement banner in which an information voice is dubbed with a performer's face and body parts to provide an effect as if the performer is providing information. The present invention relates to a system for providing an announcement advertisement banner for a website, which is inserted into a website to increase advertisement effect.

An advertising banner is a simple box-type advertisement that generally occupies a part of a web page, and when a user clicks on it, a more in-depth advertisement is displayed or access to an advertiser's website is provided.

Such an advertisement banner is a very important factor contributing to an increase in product sales online. Recently, advertisement banners are widely used as means for promoting products and services not only on web portal pages but also on pages such as instant messaging applications and social network service (SNS) applications.

Since the advertisement banner should easily catch the user's eye while scrolling the page or moving to another page, it is important to design the product so that the characteristics of the product are appropriately reflected. In addition, there is a need to design various advertisement banners in which images to be displayed on the advertisement banners are appropriately arranged.

As a prior art for such advertising banners, 'online banner advertisement display method and system' is disclosed in Korean Patent Publication No. 10-2013-0040393.

The method of displaying an online banner advertisement according to the prior art includes controlling a display of a list of contents available on a site in a first area, which is a part of a screen connected to a site for using a content providing service; Controlling to display a list of banner contents including contents related to available contents on a second area, which is a part of the access screen excluding the first area; and controlling display of banner content corresponding to list information selected by an input means of a user terminal among list information on at least one banner content included in the banner content list in the first area. to be

According to the prior art, the banner content related to the content is displayed where the list of content available to users is displayed, so that the banner content can be intuitively recognized by users without consuming unnecessary screen space. there is. Through this, there is an effect of maximizing the advertising effect of the banner content for users.

However, the above-described prior art simply displays banner content in the content list, which may hinder the identification of advertising content, and has a problem that the content to be advertised is not conveyed in detail, and is applied to websites with different designs. It has limitations that make it difficult to become.

Therefore, in order to solve the above-mentioned problems, it is possible to increase the advertising effect by providing the contents of the advertisement in the form of voice. There is a need to develop an advertisement banner providing system.

Korean Patent Publication No. 10-2013-0040393

A main object of the present invention is to increase the distinctiveness of the advertisement banner by generating and providing an announcement banner on one side of a web page in which an announcer delivers advertisement contents together with voice.

Another object of the present invention is to make a website and an announcement banner more naturally harmonize.

Another object of the present invention is to control the shape and length of the arm of the announcement banner according to the composition of the website.

In order to achieve the above object, a system for providing an announcement advertisement banner for a website according to the present invention includes a video input unit for receiving an image including a performer's face, and an audio input unit for receiving information and audio including guidance information; a banner generating module including a background removing unit that removes the background from the image, and a banner generating unit that creates an announcing banner by dubbing the informational sound to the image from which the background has been removed; a position setting module for setting an output position of the announcement banner on a web site; and a banner exposure module for displaying and reproducing the announcement banner at the output position.

Furthermore, the image includes a body part including a performer's face and an upper body including arms and hands, and the system includes an image DB storing a plurality of input images, and an arm region and a hand region in the image. a hand region detection unit that detects a hand shape object, and a hand shape object extractor that extracts and stores hand shape objects for each shape pointing in a specific direction with respect to the arm region and the hand region detected in each of the plurality of images stored in the image DB; and , a concavo-convex detection unit that detects concavo-convex including the protrusion direction of the line closest to the output position, and synthesizes the hand-shaped object with the arm region and the hand region of the image from which the background has been removed so as to point to the protrusion direction. and a synthesizing module including a hand synthesis unit generating an image, wherein the banner generation unit generates an announcing banner by dubbing the information voice to the synthesized image.

In addition, the synthesizing module may include a length control unit that controls the length of the arm region of the synthesized image according to the curvature of the irregularities.

According to the system for providing an announcement advertisement banner for a website of the present invention,

1) By providing advertisements by displaying and processing an announcement banner on a website in which a performer such as an announcer seems to provide advertisements in the form of audio, the advertisement effect for advertisement viewers who may be website users with the announcement banner inserted is increased,

2) By changing the shape of the arm and hand of the announcement banner according to the concavity and convexity of the line closest to the display position of the announcement banner, the announcement banner can show the appearance of pointing and emphasizing a specific direction of the website, harmonizing more naturally with the composition of the website At the same time, you can provide an announcement banner highlighting a specific part of the website,

3) It is characterized by realizing a more natural harmony of composition through adjusting the arm length of the announcement banner.

1 is a conceptual diagram of the system of the present invention;
Figure 2 is a block diagram showing the detailed configuration of the system of the present invention.
3 is a conceptual diagram of a website on which an announcement banner according to the present invention is displayed;
4 is a conceptual diagram showing an announcement banner in which the color of the present invention is changed;
5 is a conceptual diagram illustrating an example of extracting a face region and a mouth region;
6 is a conceptual diagram illustrating an example of extracting an arm region and a hand region;
7 is a conceptual diagram illustrating a synthesis example indicating a protruding direction;

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. The accompanying drawings are not drawn to scale, and like reference numbers in each drawing indicate like elements.

1 is a conceptual diagram showing a schematic configuration of the system of the present invention.

Referring to FIG. 1, the system for providing an announcement advertisement banner for a website of the present invention includes a performer 10 and a central control server 20.

The performer (10) is a person who appears on the announcement banner (30) and conveys the contents of the advertisement, and is a company representative, an expert in the field, an influencer, an AI human, an announcer or anchor, a weather forecaster, and a show host for home shopping. etc. can be the performer 10. The system of the present invention takes a video of such a performer 10, produces an announcing banner 30 based on the video taken of the performer 10 in the central control server 20 to be described later, and announces the banner on the website. (30) is displayed.

The video taken of the performer 10 may basically include a part of the upper body including the face, or a part of the body including parts of the upper and lower body including the face of the performer 10. Accordingly, the entire body or body part of the performer 10 may be captured as an image.

Furthermore, the preferably created announcement banner 30 includes an informational voice, which is a voice related to the advertising content, so as to provide advertisements more effectively.

The central control server 20 removes the background from the input image, creates an announcement banner 30 by dubbing the information voice to the image from which the background has been removed, and displays the announcement banner 30 on the website screen. to provide. That is, the subject of the separate announcement advertisement banner providing system of the present invention can be said to be the central control server 20 .

The central control server 20 is a series of subjects for realizing the system of the present invention, and includes a server PC and a network communication network. In addition, the central control server 20 has a central processing unit (CPU) and a program that can be executed in the central processing unit on a hardware basis having storage means such as memory and hard disk, that is, software is installed and can run this software. A series of detailed configurations of such software will be described later as units such as 'modules', 'units' and 'parts'.

Configurations such as 'modules' or 'units' or 'interfaces' or 'parts' are installed and stored in the storage means of the central control server 20, such as software or FPGA or ASIC executed via CPU and memory. Means a piece of hardware. At this time, the configuration of 'module' or 'unit' or 'interface' is not limited to hardware, and may be configured to be in an addressable storage medium or configured to reproduce one or more processors. As an example, 'module' or 'part' or 'interface' refers to components such as software components, object-oriented software components, class components, and task components, processes, functions, and attributes. fields, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables.

Functions provided by these 'modules' or 'units' or 'interfaces' may be combined into a smaller number of components and 'units' or 'modules', or may be combined into additional components and 'units' or 'modules'. can be further separated.

In addition, the central control server 20 means all kinds of hardware devices including at least one processor, and can be understood as encompassing software configurations operating in the hardware device according to the embodiment. For example, a computing device as an example of a server may be understood as including all of smart phones, tablet PCs, desktops, laptops, and user clients and applications running on each device, but is not limited thereto.

Hereinafter, based on the configuration of the central control server 20, the detailed configuration of the system for providing an announcement advertisement banner for a website according to the present invention will be described with drawings.

2 is a block diagram showing the detailed configuration of the system of the present invention, and FIG. 3 is a conceptual diagram of a website on which an announcement banner according to the present invention is displayed.

Referring to FIGS. 2 and 3, the system for providing an announcement advertisement banner for a website of the present invention is characterized in that it includes a banner creation module 100, a location setting module 200, and a banner exposure module 300. .

The banner generation module 100 performs a function of generating an announcement banner 30 by removing a background from an image including the face of the performer 10 and dubbing an informational voice. To this end, the banner generation module 100 It includes a video input unit 110, an audio input unit 120, a background removal unit 130, and a banner generator 140.

The image input unit 110 performs a function of receiving an image including the face of the performer 10 as an input. Here, the image refers to a photograph of the face of the performer 10, and furthermore, a part of the body including the face and the upper body wearing the costume, preferably against the background of the chroma key screen, while the performer 10 gestures, information It may be an image that conveys.

At this time, in order to convey information while making a gesture, the image of the present invention may be a photograph of a part of the body including the upper body including the hands as well as the face of the performer 10 .

Therefore, such at least one image is input through the image input unit 110, and here there is no limit to the number of images input through the image input unit 110, and the more images there are, the easier it is to extract objects for synthesis, which will be described later. The number of images input through the image input unit 110 may be tens, hundreds, thousands, or even tens of thousands. In order to more realistically synthesize facial expressions or mouth shapes in a configuration to be described later, it is preferable that the number of images input through the image input unit 110 be a number capable of forming big data.

There are no restrictions on where the video is filmed, so there are no restrictions on studio shooting, outdoor shooting, set shooting, etc. Therefore, images taken at various places and times may be input, but preferably background removal is facilitated by inputting images of parts of the body including the face and upper body of the performer 10 against the background of the chroma key screen. can do.

The voice input unit 120 performs a function of receiving an information voice including guide information. Here, the guide information is the content of information to be provided. For example, when a specific service is to be advertised through the announcement banner 30, the content of the advertisement becomes guide information, and the voice information can be a voice that reads the guide information.

At this time, the person providing the information voice may preferably be the performer 10, but it is also possible for someone other than the performer 10 to provide the information voice, and in the case of the information voice, it is preferably input with clear pronunciation. It is desirable to increase discrimination. Alternatively, it is also possible to input guide information as text, convert it into voice through a TTS (text to sound) function, and input the information as voice.

The background removal unit 130 performs a function of removing the background from the image. Here, a separate limitation is not imposed on the background removal method, but recognizes an object in the image, that is, the performer 10, and other objects except for the area of the object. The background may be removed by deleting the area.

More specifically, the performer area including the performer 10 and the background area in the video are detected and the background area is removed. If the entire body of the performer 10 is captured in the video, the entire body is captured. becomes a performer area, and when a body part is photographed, the body part becomes a performer area. In other words, the performer area refers to the entire body or part of the performer 10 .

In this case, in order to detect the performer area and the background, the outline of the person (performer area) captured in the video is detected, the area within the closed curve formed by the outline is detected as the performer area, and the area not detected as the performer area is detected as the background. can be detected. Since this can be done by applying a conventional method for detecting a person in an image, a separate description will be omitted.

Furthermore, as described above, when an image has a chroma-key screen as a background, background removal may be easier, which can be implemented using a conventional chroma-key technique, and a performer area is detected In this case, it can be said that the background except for the detected performer region is removed to make a blank, and this is done for each frame, and finally all the backgrounds can be removed from the chroma key image.

The banner generation unit 140 dubs the information voice input through the voice input unit 120 to the image from which the background has been removed through the background removal unit 130, that is, the image in which only the performer area remains, and announces the banner ( 30) (announcing banner).

Here, the announcement banner 30 refers to a banner that gives a feeling that the performer 10, that is, the performer 10, which can also be called the host of the advertisement, provides informational voice, that is, the announcer delivers the advertisement. refers to

Therefore, when a sound such as a voice is inserted into the image from which the background has been removed, the corresponding sound, that is, the sound is muted and the information voice is dubbed so that the image with the background removed and the information voice are displayed through the announcing banner 30. to be provided together.

The location setting module 200 performs a function of setting the output location of the created announcement banner 30 on the website.

Here, the web site may also refer to a web page, and therefore, when an output position is designated on the web page, the announcement banner 30 may be moved when scrolling up or down. In this case, the output location can be set as coordinates on the web page. In other words, depending on the web page, only all or part of the announcement banner 30 may be output to advertisement viewers according to the conversion or movement of the web page.

Alternatively, the output location may be designated such that the announcement banner 30 is located at specific coordinates on the screen where the website is displayed. In other words, even if the web page is scrolled, the output position can be designated so that the announcement banner 30 is output at a specific coordinate of the display screen on which the web site is output so that the announcement banner 30 is output at a specific position on the web site screen. In this case, the output location may be set as coordinates on the screen. In other words, if the output position is designated as one side of the coordinates on the screen, the announcement banner 30 is not subordinate to the web page, so even if the page scrolls or the web page moves, the announcement banner 30 is displayed at a fixed position. .

Most preferably, the position setting module 200 sets the output position of the announcement banner 30 on the screen on which the web site is output, so that the announcement banner 30 is not dependent on the web page, so scrolling of the page occurs or web Even if the page is moved, it may be preferable to output the announcement banner 30 at a fixed location to increase the advertising effect.

The banner exposure module 300 performs a function of displaying and reproducing the announcement banner 30 at a set output location. At this time, since the announcement banner 30 is in the form of a video provided by dubbing the information voice to the video of the performer 10, when displaying and playing the announcement banner 30, a separate play/stop button is added to display the announcement banner ( 30) can be played or not reproduced, of course.

Therefore, according to the system for providing an announcement advertisement banner of the present invention, the announcer 10 displays the announcement banner 30 in the form of voice on the website and provides the advertisement, thereby providing the corresponding announcement banner 30. (30) has an effect of increasing the advertisement effect for advertisement viewers who may be web site users inserted.

4 is a conceptual diagram showing an announcement banner with a changed color according to the present invention.

Referring to FIG. 4 , in the above description, the image of the present invention may include a part of the body including the face of the performer 10 and the upper body wearing the costume. Furthermore, in this case, the announcement banner may also include a part of the body including the upper body wearing the background and the costume of the performer 10 as a matter of course.

At this time, if the color of the output location of the website matches the color of the clothes of the performer 10 included in the announcement banner, there is a concern that the advertisement delivery effect may be hindered due to a decrease in discrimination of the announcement banner.

Therefore, in order to increase the discrimination of the announcement banner, the system of the present invention can add a configuration for changing the color of the clothing 40 included in the announcing banner. To this end, the system may include a clothing change module 400 Preferably, the clothing change module 400 may include a color detection unit 410 and a color change unit 420.

The color detection unit 410 included in the costume change module 400 performs a function of determining the color of a designated output position. Here, if the output location is a predetermined location of the website (web page), that is, coordinates on the web page, the color of the website coordinates is identified, and the output location is a predetermined location on the screen where the website is output, that is, the coordinates of the screen. In this case, the color currently displayed at the corresponding coordinates on the screen is identified.

The color changing unit 420 performs a function of changing the color of the clothing 40 included in the announcement banner to a color that does not match the color of the output position. In other words, the color of the clothing 40 included in the announcement banner is changed to a different color from the color of the output position so that the color of the clothing 40 included in the announcement banner is different from the color of the output position.

At this time, the color of the clothing 40 may be changed to a color complementary to the color of the output position, but if the color does not match, it is also possible to change the color of the clothing 40 within a natural color range.

Therefore, through this configuration, the color of the costume 40 of the performer 10 included in the output location and the announcement banner is the same or the color of the costume 40 can be changed to increase the discrimination of the announcement banner, maximizing the advertising effect. can do.

5 is a conceptual diagram illustrating an example of extracting a face area and a mouth area.

Referring to FIG. 5 , in order to produce an announcement banner whose mouth shape is moved to match the information voice, a burden of inputting an image having the same mouth shape as the information voice is incurred every time. However, if the mouth shape of the above-described background-removed image can be changed to be the same as that of the information voice, the inconvenience of having to receive an image every time for each information voice can be resolved.

Therefore, in order to solve this problem, the system of the present invention may include a synthesis module 500, and the synthesis module 500 preferably includes an image DB 510, a face area detection unit 520, a mouth shape object extraction unit ( 530), a pronunciation analysis unit 540, and a mouth synthesis unit 550.

The image DB 510 performs a function of storing a plurality of images input through the image input unit 110 . In other words, it can be referred to as a database for images input through the image input unit 110 .

In the above description, it has been said that the number of images input through the image input unit 110 can be tens, hundreds, thousands, or even tens of thousands. More preferably, the number of images input through the image input unit 110 is big data. Since is referred to as a number that can be achieved, it can be said that such a plurality of images are stored in the image DB 510.

The face area detection unit 520 performs a function of detecting the face area 50 in the image and detecting the mouth area 60 from the face area 50 . At this time, the detection method of the face area 50 and the mouth area 60 is not limited, but preferably, the outline of the face is detected and the area within the closed curve formed by the outline is detected as the face area 50, and the face area 50 ), it is also possible to detect the inner region formed by the outline of the lips as the mouth region 60 by detecting the outline of the lips.

Here, since the image means that the face of the performer 10 is photographed and includes the face, only the face region 50 of the performer 10 is detected from the image, and the mouth region showing the mouth shape in the face region 50 It performs the function of detecting only (60).

The mouth shape object extraction unit 530 performs a function of extracting and storing a mouth shape object for each pronunciation of a voice from each of a plurality of images stored in the image DB.

For example, since each video is a video for delivering guide information about advertisement contents, the performer 10 delivers the guide information based on voice, and at this time, the mouth area 60 moves according to the pronunciation of the voice. For example, if the guide information is 'hello', pronunciation corresponding to the text constituting the guide information, such as 'an', 'nyeong', 'ha', 'se', and 'yo' is made.

At this time, the shape of the mouth is different for each pronunciation of each letter, and the mouth shape object extraction unit 530 extracts and stores the mouth shape object for each pronunciation. That is, only the mouth region 60 is detected and cut out from each image, and only the mouth region 60 is stored, and an object in which the shape of the mouth is divided for each pronunciation is stored. For example, a part of an image in which only the mouth region 60 corresponding to the shape of the mouth pronouncing 'an' in the image is cut out, that is, a part in which only the mouth region 60 is edited becomes an object for the pronunciation of 'an'. will be.

At this time, the mouth shape object for each pronunciation may preferably be stored for each Korean pronunciation, and furthermore, in a foreign language, only the mouth region 60 is cut and stored in the image for each pronunciation of the corresponding language and stored as an object for the mouth shape for each pronunciation. possible.

Furthermore, since there is a standard pronunciation method in Korean, the contents of the information to be provided in the input video are converted into text, and then written according to the pronunciation according to the standard pronunciation method, and then divided into syllables, and the mouth area of the video corresponding to the syllable It is also possible to cut only (60), extract it as an object for the mouth shape for each pronunciation, and store it.

The pronunciation analysis unit 540 performs a function of analyzing the pronunciation of the information voice, and analyzes the pronunciation of the information voice corresponding to the guide information as much as the actual pronunciation may be different from the text corresponding to the guide information.

For example, as described above, if 'tomorrow's weather will be clear' is information information, the pronunciation of the information voice is 'tomorrow's weather is mal? At this time, the pronunciation analysis of the information voice is preferably based on the standard pronunciation method.

According to the pronunciation of the information voice analyzed by the pronunciation analysis unit 540, the mouth synthesis unit 550 synthesizes a mouth-shaped object corresponding to the pronunciation to the mouth region 60 of the image from which the background has been removed, and processes the synthesized image. generate

That is, according to the pronunciation and playback speed of the information voice, the mouth shape object corresponding to the pronunciation is synthesized with the mouth area 60 detected in the background-removed image, which is the source of synthesis, and the mouth shape in the background-removed image. This is to create the altered composite image.

As described above, when 'the weather will be clear tomorrow' is information information, the pronunciation of the information voice is 'what is the weather tomorrow? The ''day'' of ''day'' is a mouth-shaped object that pronounces each syllable for ''word''? Synthesis processing is performed on the mouth area 60, and a mouth-shaped object is synthesized to correspond to the contents and pronunciation of the corresponding information voice, the reproduction speed of the information voice, and the length of the voice.

Furthermore, when a synthesized video is generated in this way, the banner generator 140 completes an announcement banner by dubbing the synthesized video with informational voice, which is the basis of the analysis by the pronunciation analysis unit 540.

According to this, as the announcer 10 included in the actual announcement banner pronounces the information voice, the mouth shape of the announcement banner and the pronunciation of the information voice match, so that a more natural announcement banner can be provided, and furthermore, all information voices Even if an image is not taken, an announcement banner can be produced by inserting the extracted object into the image, thereby reducing the production cost of the announcement banner.

In addition, in order to create a more realistic composite image, the boundary surface smudging at the boundary surface of the mouth area 60 where the object for the mouth shape is inserted, that is, the boundary surface smearing process is performed so that the synthesized part is not visible. It is also possible to do To this end, the synthesis module 500 may further include a skin pixel detection unit 560 and a correction processing unit 570 .

The skin pixel detection unit 560 detects a boundary between the mouth region 60 and the face region 50 in the generated composite image, that is, a skin pixel at the boundary corresponding to the outline of the synthesized mouth region 60 . Here, the skin pixel refers to detecting a pixel representing a skin color, that is, a general skin color ranging from ivory to apricot, and furthermore to brown. It is to detect a pixel having .

The correction processor 570 performs a function of smudging the detected boundary area and correcting saturation and brightness.

Here, the smudge of the boundary part means blurring the boundary by blurring the synthesized boundary surface by processing the boundary part to be smudged, and making the boundary of the synthesized part unrecognizable. Through this, the skin of the face area 50 can look cleaner, and at the same time, a more natural composite image can be provided as the boundary of the synthesized part becomes difficult to recognize because the boundary is blurred.

Furthermore, correcting the saturation and brightness can be said to be correcting the color and brightness of the synthesized image. Therefore, objects for the shape of the mouth are synthesized to occur between the face area 50 and the mouth area 60. A natural boundary surface can be created by lowering the possible difference in saturation (color) and brightness (brightness) and evening the tone in the skin pixels in the boundary area.

Therefore, according to the configuration of the skin pixel detection unit 560 and the correction processing unit 570, an announcement banner is produced by synthesizing and processing mouth shape objects without generating images for all information voices. There is an effect of providing a more natural announcement banner by minimizing the awkwardness of the boundary between the mouth area 60 and the face area 50.

6 is a conceptual diagram illustrating an example of extracting an arm region and a hand region, and FIG. 7 is a conceptual diagram illustrating a synthesis example indicating a protruding direction.

In the above description, it is said that the announcement banner of the present invention may provide information about advertisement contents while the performer 10 makes a gesture. In this case, in order to deliver information while making a gesture, the video of the present invention It was said that it may include not only the face of the performer 10 but also body parts including the upper body including the hands. Here, the inclusion of the hand may mean that the arm and the hand connected to the arm are included among the upper body.

Referring to FIGS. 6 and 7 , when a hand is included, it is possible to manufacture an announcing banner pointing in a specific direction according to a change in hand shape. ), a hand shape object extraction unit 590, a concavo-convex detection unit 600, and a hand synthesis unit 610 may be further included.

The hand region detection unit 580 performs a function of detecting the arm region 70 and the hand region 80 from the plurality of images stored in the above image DB. Here, the detection of the hand region 80 can be referred to as detecting the arm region 70 in the image and detecting the hand region 80 connected to the end of the arm region 70 . This may be referred to as detecting an outline having a shape corresponding to an arm and a hand, and detecting an area within a closed curve formed by the outline as an arm area 70 and a hand area 80.

Therefore, the arm region 70 refers to a starting point connected from the shoulder to an end point connected to the hand, and it can be said that the hand region 80 is from the end point to the end of the arm region 70 .

The hand shape object extraction unit 590 performs a function of extracting and storing hand shape objects for each shape pointing in a specific direction with respect to the hand region 80 detected in each of a plurality of images stored in the image DB 510. Here, the hand-shaped objects may be extracted for each shape pointing in a specific direction, such as left, right, up, and down.

For example, a hand pointing left with an arm extending to the left becomes a hand object object pointing left, and a hand pointing upward with an arm extending upward and a hand pointing upward. It can be a shape object.

The concavo-convex detection unit 600 performs a concavo-convex function including a protruding direction of a line located in a region adjacent to an output position. Here, since the output location may be a predetermined location of the website or a predetermined location of the screen on which the website is output, a specific line included in the website may be included in an area near the corresponding output location.

Therefore, if the irregularities including the projection direction of the output position and the line closest to the output position are identified in the area near the output position, in particular, how many irregularities are present in the output position and the line closest to the output position of the announcement banner, and furthermore, in what direction it is protruded. can figure out For example, if it protrudes in the left direction and protrudes by 10 pixels, the protrusion direction may be left and the degree of concavo-convex may be 10 pixels. In this way, protruding directions and irregularities can be identified.

If the line closest to the output position is a simple straight line (vertical or horizontal line), it can be determined that there is no protruding direction and degree of unevenness.

The hand synthesizer 610 performs a function of generating a synthesized image by synthesizing a hand-shaped object with the arm region 70 and the hand region 80 of the image from which the background has been removed so as to point in the protruding direction.

For example, when the extrusion direction is left, a hand-shaped object pointing to the left, that is, a hand-shaped object in which both the arm region 70 and the hand region 80 point to the left are converted into the arm region 70 of the image from which the background is removed. and synthesized with the hand region 80 to generate a synthesized image.

As another example, if the protrusion direction is right, the hand-shaped object pointing to the right, that is, the arm region 70 and the hand region 80 both point to the right, the arm region 70 of the image from which the background is removed. and synthesized with the hand region 80 to generate a synthesized image.

When the extrusion direction is upward, a hand-shaped object pointing upward, that is, a hand-shaped object in which both the arm region 70 and the hand region 80 point upward, is displayed as the arm region 70 and the hand region of the image from which the background has been removed. (80) to generate a synthesized image.

If the extrusion direction is downward or nonexistent, that is, if the line closest to the output position is a straight line with no extrusion direction, the downward-pointing hand-shaped object, that is, the arm area 70 and the hand area 80 both point to the right. A synthesized image in which the performer 10 appears to be at attention may be generated by combining the pointing hand-shaped object with the arm region 70 and the hand region 80 of the image from which the background is removed.

Furthermore, when a synthesized video is generated in this way, the banner generating unit 140 dubs information voice over the generated synthesized video to generate an announcing banner, and the generated announcing banner can be displayed and reproduced in an output position. .

Therefore, through this, in the announcement banner displayed on the website, the shape of the arm and hand of the announcement banner is modified according to the concavity and convexity of the line adjacent to the display position of the announcement banner, thereby indicating the appearance of the announcement banner pointing in a specific direction of the website and emphasizing it. Therefore, it is of course possible to provide an announcing banner that more naturally blends with the composition of the website and to provide an emphatic effect for the website.

Furthermore, it is possible to vary the length of the arm region 70 included in the synthesized image according to the above-described curvature of the concavo-convex so as to harmonize with the web site more naturally. To this end, the synthesis module 500 may include a length controller 620.

Furthermore, the length of the arm region 70 here means the length of the longest part of the calculated arm region 70 . That is, the length of the longest axis in the region identified as the arm region 70 becomes the length of the arm region 70 .

The length control unit 620 provides a function of varying the length of the arm region 70 included in the composite image according to the degree of curvature of the concavo-convex. Here, preferably, since a higher degree of curvature means a steeper and more abrupt protrusion, the arm region ( 70) may be adjusted to be longer so that the protruding direction is naturally directed. In addition, the lower the degree of curvature, the more gently protrudes, meaning that there are almost no irregularities, so that the length of the arm region 70 is hardly changed to prevent the line from being invaded.

Here, the variable length of the arm region 70 means that the length of the arm region 70 is increased or decreased. Therefore, the length of the arm region 70 is reduced or increased by changing the size of a specific object on a video or image. it is possible Since this can be implemented through a conventional photo correction program or application, a detailed description thereof will be omitted.

At this time, more preferably, the above-described length control unit may vary the length of the arm region 70 through Equation 1.

Equation 1,

(Where A' is the length of the arm region, A is the length of the arm region, r is the radius of curvature (pixel), and D is the shortest distance between the end point of the hand region and the line (pixel))

For example, if the length of the arm area is 100 pixels, the radius of curvature is 80 pixels, and the shortest distance between the end point of the hand area and the line is 10 pixels,

로 산출된다.

is calculated as

Therefore, the length of the variable arm region shows a distribution of values that become smaller as the radius of curvature increases (as the degree of curvature decreases) and becomes longer as the radius of curvature decreases (as the degree of curvature increases).

At this time, rather than simply performing linear correction, the distance of the arm region can be varied by comparing the radius of curvature and the shortest distance between the end point of the hand region and the line through the hypertangent function, so that the maximum value of the hypertangent The multiplied value, i.e., a value exceeding 0.3 was prevented from appearing. Therefore, the length of the arm region is not varied by more than 30%, preventing unnatural length adjustment, and reflecting a correction value obtained by taking a hypertangent rather than an abrupt linear correction, thereby enabling a more natural and gradual correction.

Therefore, according to the variation of the length of the arm region 70 through the hypertangent function, the announcement banner 30 can be obtained by varying the length of the arm region 70 included in the synthesized image according to the curvature of the unevenness of the line included in the website. The protruding direction of the line included in the website can be naturally directed so that it blends more naturally with the composition of the website. At this time, the length of the arm area 70 included in the announcement banner 30 is unnaturally varied Furthermore, even in the change of the music of the web page, the variation is gradually made to provide a feature that enables a more natural variation.

With this configuration, the arm region 70 is adjusted according to the curvature of the nearest line of the announcing banner, so that the curved line and the arm and hand of the announcing banner are more naturally harmonized, and at the same time, the effect of enhancing the emphasis is achieved.

As described above, although the configuration and operation of the system for providing an announcement advertisement banner for a website according to the present invention is expressed in the above description and drawings, this is merely an example and the spirit of the present invention is limited to the above description and drawings. It is not, and of course, various changes and changes are possible within a range that does not deviate from the technical spirit of the present invention.

10: performer 20: central control server
30: Announcement Banner 40: Costume
50: face area 60: mouth area
70: arm area 80: hand area
100: banner generation module 110: video input unit
120: voice input unit 130: background removal unit
140: banner generation unit 200: location setting module
300: Banner exposure module 400: Costume change module
410: color identification unit 420: color change unit
500: synthesis module 510: image DB
520: face area detection unit 530: mouth shape object extraction unit
540: pronunciation analysis unit 550: mouth synthesis unit
560: skin pixel detection unit 570: correction processing unit
580: hand area detection unit 590: hand shape object extraction unit
600: concavo-convex part 610: hand composite part
620: length control

Claims (8)

As an announcement advertisement banner providing system for a website,
An image input unit that receives an image including a part of the body including the face of the performer and the upper body including arms and hands, an audio input unit that receives information audio including guide information, and a background removal unit that removes the background from the image and a banner generating module including a banner generating unit generating an announcing banner by dubbing the informational voice to the image from which the background has been removed;
a position setting module for setting an output position of the announcement banner on a web site;
An image DB storing a plurality of input images, a hand region detector detecting an arm region and a hand region from the images, and a specific direction for the arm region and the hand region detected in each of the images stored in the image DB. A hand-shaped object extraction unit for extracting and storing hand-shaped objects for each pointed shape, a concavo-convex detection unit for identifying concavo-convex and convexities including the protrusion direction of a line closest to the output position, and pointing to the protrusion direction. A hand synthesis unit generating a synthesized image by synthesizing the hand-shaped object with the arm region and the hand region of the image from which the background has been removed; and a length controller varying the length of the arm region of the synthesized image according to the curvature of the concavo-convex curve. A synthesis module comprising;
A banner exposure module for displaying and reproducing the announcement banner at the output location;
The banner generator,
dubbing the informational voice to the synthesized video to generate an announcing banner;
The length controller,
A system for providing an announcement advertisement banner, characterized in that the length of the arm region is varied through Equation 1 below.
Equation 1,

(Where A' is the length of the arm region, A is the length of the arm region, r is the radius of curvature (pixel), and D is the shortest distance between the end point of the hand region and the line (pixel)) According to claim 1,
The location setting module,
The system for providing an announcement advertisement banner, characterized in that setting an output position of the announcement banner on a screen on which the web site is displayed. According to claim 1,
The video,
It includes parts of the body, including the face of the performer and the upper body wearing the costume,
The system,
a color grasping unit for grasping the color of the output position;
and an apparel change module including a color changer for changing the color of the apparel included in the announcement banner to a color that does not match the color of the output position. According to claim 1,
The synthesis module,
a face area detection unit that detects a face area from the image and detects a mouth area from the face area;
a mouth shape object extraction unit extracting and storing a mouth shape object for each speech pronunciation from the mouth region detected in each of the images stored in the image DB;
a pronunciation analyzer for analyzing the pronunciation of the information voice;
Further comprising a mouth synthesis unit for generating a synthesized image by synthesizing the mouth-shaped object with the mouth region of the image from which the background has been removed according to the pronunciation,
The banner generator,
The system for providing an announcement advertisement banner, characterized in that an announcement banner is generated by dubbing the information voice to the synthesized video. According to claim 4,
The synthesis module,
a skin pixel detector detecting skin pixels at a boundary between the mouth region and the face region belonging to the synthesized image;
and a correction processing unit for smudging the boundary and correcting saturation and brightness. delete delete delete

Images