WO2021172633A1

WO2021172633A1 - Method and device for recognizing moving image content, and image processing system including same

Info

Publication number: WO2021172633A1
Application number: PCT/KR2020/002884
Authority: WO
Inventors: 이상민; 김형진; 조청호
Original assignee: (주)뉴빌리티
Priority date: 2020-02-28
Filing date: 2020-02-28
Publication date: 2021-09-02

Abstract

The present invention relates to a method and device for recognizing moving image content, and an image processing system including same. The device comprises: a storage unit for storing at least one recognition module which recognizes information included in moving image content; and a control unit which designates moving image content in which information is to be recognized, and connects, in a layered formed, at least one area of the moving image content and the at least one recognition module stored in the storage unit, and thereby recognizes the information included in the moving image content through the at least one recognition module. [Representative drawing] Figure 1

Description

Video content recognition method and apparatus, and image processing system including the same

The present invention relates to image processing, and more particularly, to a method and apparatus for recognizing video content, and an image processing system including the same.

A moving picture, such as a game, includes various information including letters, numbers, and images. If information is automatically extracted from these videos over time, or information about the section or location in which each information appears, it is convenient to provide additional information related to the information, and it can be effectively used to provide various application services. have.

However, existing image recognition programs or game data recognition programs require development capacity such as Open Source Computer Vision (Open CV) or machine learning. In addition, in order to analyze game data, it was necessary to produce a large amount of one-time programs such as analysis model selection, tuning data production program, etc., which required a lot of repetitive work.

[Prior art literature]

(Patent Document 1) Domestic Patent Application No. 10-2005-0044005

(Patent Document 2) Domestic Registered Patent Publication No. 10-1104699

The present specification has been devised to solve the above problems, and even if not a developer, a plurality of users through a graphical user interface environment can easily tune and produce a program for recognizing information included in video content. An object of the present invention is to provide a method and an apparatus, and an image processing system including the same.

Another object of the present specification is to provide a video content recognition method and apparatus capable of providing a simplified user environment such as an automated image recognition workflow, automatic creation of a screen data collection program, and automation of model data tuning through the web, and includes the same An image processing system is provided.

According to an embodiment of the present specification, an image processing system according to the present specification includes: a web server for storing a plurality of recognition modules for recognizing information included in video content; and receiving at least one recognition module among the plurality of recognition modules from the web server, designating the video content for which the information is to be recognized, and layering at least one region of the video content and the at least one recognition module and a video content recognition device for recognizing information included in the video content through the at least one recognition module by being connected to the video content.

Preferably, the apparatus further comprises a plurality of cross-validation devices for inputting various types of information included in the video content into a recognition module, and performing cross-validation on the result values of the recognition module to update the plurality of recognition modules. characterized in that

According to another embodiment of the present specification, an apparatus for recognizing video content according to the present specification includes: a storage unit configured to store at least one recognition module for recognizing information included in video content; and designating video content for which information is to be recognized, and connecting at least one area of the video content and the at least one recognition module stored in the storage unit in a layer form to obtain the video content through the at least one recognition module It includes a control unit for recognizing the included information.

Preferably, the control unit indexes the video content in a time table, sets layers for each index, and inserts the recognition module into each layer to generate the screen data collection program.

Preferably, the control unit sets a mask for the video content, sets an image resolution and a change characteristic for each resolution of each region in the mask, and then selects a layer to be used for each index.

Preferably, the control unit selects any one of linear and non-linear as a characteristic of change for each resolution of each region.

Preferably, when non-linearity is selected as a change characteristic for each resolution of each region, the control unit changes the resolution based on a coordinate value input from a user.

Preferably, the controller changes the resolution through spline interpolation for a resolution to which a coordinate value is not input.

Preferably, the layer to be used for each index is a base layer, a shake correction layer that cuts a part of an area, detects position values of internal feature points, and compares the detected position values with a default value to correct shake of the area, translucent It characterized in that it includes at least one of the layer, and an additional layer that designates a region having a different change characteristic for each position and resolution.

Preferably, the control unit designates an area to which the recognition module is input on the layer through a UI Location Filling Layer, and inserts the recognition module into the area location filling layer. .

According to another embodiment of the present specification, the method for recognizing video content according to the present specification is a method for recognizing a video content of a video content recognizing apparatus for recognizing information included in video content, and recognizing information included in the video content receiving and storing at least one recognition module; designating the video content for which the information is to be recognized; and recognizing information included in the video content through the at least one recognition module by connecting at least one region of the video content and the at least one recognition module in a layered form.

Preferably, the step of recognizing the information included in the video content includes: indexing the video content in a time table; setting a layer by index; and inserting the recognition module into each layer.

Preferably, the step of setting the layer by index comprises: setting a mask on the video content; setting an image resolution and a change characteristic for each resolution of each region in the mask; and selecting a layer to be used by index.

Preferably, the setting of the change characteristic for each resolution comprises: selecting one of linear and non-linear as the change characteristic for each resolution of each region; and changing the resolution based on a coordinate value input from a user when non-linearity is selected as a change characteristic for each resolution of each region.

Preferably, the step of inserting the recognition module comprises: designating an area to which the recognition module is input on the layer through a UI Location Filling Layer; and inserting the recognition module into the region location filling layer.

As described above, according to the present specification, a plurality of users can easily tune and produce a program for recognizing information included in video content through a graphical user interface environment, even if not a developer.

In addition, it is possible to provide a simplified user environment such as an automated image recognition workflow, automatic creation of a screen data collection program, and automation of model data tuning through the web.

In addition, by providing a capcha in the production of a program performed for each user, accuracy and efficiency can be improved.

1 is a block diagram showing a schematic configuration of an image processing system according to a first embodiment of the present invention;

2 is a block diagram showing a schematic configuration of the inside of the video content recognition apparatus according to the first embodiment of the present invention;

3 is a flowchart illustrating a video content recognition method according to a first embodiment of the present invention;

4 is a flowchart illustrating a method for generating a screen data collection program according to a first embodiment of the present invention;

5 is a flowchart illustrating a method of setting a layer for each index according to the first embodiment of the present invention;

6 is a view showing an example screen of a time table according to the first embodiment of the present invention;

7 is a view showing an example screen of an image recognition program according to the first embodiment of the present invention;

8 is a view for explaining a method of designating a region to be input to a recognition module using a region position filling layer;

9 is a diagram showing a schematic configuration of the inside of a non-verbal information delivery device according to a second embodiment of the present invention;

10 is a view showing the configuration of a solenoid module according to a second embodiment of the present invention, and

11 is a flowchart illustrating a non-verbal information delivery method according to a second embodiment of the present invention.

It should be noted that the technical terms used herein are used only to describe specific embodiments, and are not intended to limit the present invention. In addition, the technical terms used in this specification should be interpreted in the meaning generally understood by those of ordinary skill in the art to which the present invention belongs, unless otherwise defined in this specification, and excessively inclusive. It should not be construed in the meaning of a human being or in an excessively reduced meaning. In addition, when the technical terms used in the present specification are incorrect technical terms that do not accurately express the spirit of the present invention, they should be understood by being replaced with technical terms that those skilled in the art can correctly understand. In addition, general terms used in the present invention should be interpreted as defined in advance or according to the context before and after, and should not be interpreted in an excessively reduced meaning.

Also, as used herein, the singular expression includes the plural expression unless the context clearly dictates otherwise. In the present application, terms such as "consisting of" or "comprising" should not be construed as necessarily including all of the various components or various steps described in the specification, some of which components or some steps are It should be construed that it may not include, or may further include additional components or steps.

In addition, the suffixes "module" and "part" for the components used in this specification are given or mixed in consideration of the ease of writing the specification, and do not have distinct meanings or roles by themselves.

Also, terms including an ordinal number such as first, second, etc. used herein may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, a first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component.

[Explanation of code]

110: web server 120: video content recognition device

130: a plurality of cross-validation devices 210: communication unit

220: screen data collection program generation unit 230: storage unit

240: screen data collection unit

Hereinafter, a preferred embodiment according to the present invention will be described in detail with reference to the accompanying drawings, but the same or similar components are assigned the same reference numerals regardless of reference numerals, and redundant description thereof will be omitted.

In addition, in the description of the present invention, if it is determined that a detailed description of a related known technology may obscure the gist of the present invention, the detailed description thereof will be omitted. In addition, it should be noted that the accompanying drawings are only for easy understanding of the spirit of the present invention, and should not be construed as limiting the spirit of the present invention by the accompanying drawings.

1 is a block diagram showing a schematic configuration of an image processing system according to a first embodiment of the present invention.

Referring to FIG. 1 , the image processing system according to the present invention may include a web server 110 , a video content recognition device 120 , and a plurality of cross-verification devices 130 .

The web server 110 stores a plurality of recognition modules for recognizing information included in an image recognition program and video content.

In addition, the web server 110 may include module data including attribute information of each recognition module inserted by the user through a layer, and learning data including non-post-processed data and post-processed data.

Here, the non-post-processed data represents data in which information about the data is not indexed, that is, data that is not optimized with a recognition module. In addition, the post-processing data represents data obtained by indexing each data and result values of a recognition module that a user or others manually want to optimize. That is, the post-processing data represents data optimized with the recognition module through cross-validation of a plurality of cross-validation devices 130 to be described later. To this end, the web server 110 may provide open sources for the plurality of recognition modules to the plurality of cross-validation devices 130 .

As such, the web server 110 may provide an updated authentication module based on the learning data cross-verified by the plurality of cross-validation devices 130 to the video content recognition device 120 .

The moving image content recognition apparatus 120 accesses the web server 110 and receives at least one recognition module for recognizing an image recognition program and information included in the moving image content from the web server 110 .

The video content recognizing apparatus 120 designates video content for which information is to be recognized on an Operating System (OA) through the received image recognition program, and includes at least one area and at least one area for recognizing information in the video content. A screen data collection program is created by connecting the recognition modules of A detailed structure and operation of the video content recognizing apparatus 120 will be described with reference to FIG. 2 .

The plurality of cross-verification devices 130 input various information included in the video content to the recognition module based on the open source for the plurality of recognition modules provided by the web server 110, A plurality of recognition modules are updated by performing cross-validation. Specifically, one or a small group of users initially creates some data and measures the accuracy of the recognition module to create a standard, and a plurality of verification personnel perform verification on each data to agree more than a preset number of people In this case, it is judged as valid data.

In addition, the plurality of cross-validation devices 130 may improve the accuracy of the recognition module through an automatic input prevention system such as a capcha. For example, the plurality of cross-validation authentication devices 130 may cause CAPTCHAs to appear frequently in the case of a recognition module requiring additional data acquisition in consideration of the number of users and satisfaction of the recognition module.

In the embodiment of the present invention, the moving image content recognition apparatus 110 and the plurality of cross-verification apparatuses 130 are separately described, but the moving image content recognition apparatus 110 may be each cross-verification apparatus 130 .

2 is a block diagram showing a schematic configuration of the inside of the apparatus for recognizing video content according to the first embodiment of the present invention.

Referring to FIG. 2 , the moving image content recognition apparatus 110 according to the present invention may include a communication unit 210 , a screen data collection program generation unit 220 , a storage unit 230 , and a screen data collection unit 240 . can Here, the screen data collection program generation unit 220 and the screen data collection unit 240 constitute the control unit.

The communication unit 210 transmits and receives data to and from the web server 110 through wired/wireless communication. That is, the communication unit 210 receives an image recognition program, at least one recognition module, and various data related to each recognition module from the web server 110 , and a video content recognition process or screen data collection process with the web server 110 . It is possible to transmit various data generated in the web server (110).

The screen data collection program generating unit 220 designates video content for which information is to be recognized on the operating system system through the image recognition program, and forms at least one region and at least one recognition module for recognizing information in the video content in a layered form. to create a screen data collection program.

Specifically, in order to generate the screen data collection program, the screen data collection program generation unit 220 first calls the video content, and uses an image editing tool from the beginning to the end of the video content to be recognized. cut Then, the screen data collection program generating unit 220 indexes the video content in the time table after editing the beginning and the end of the video content. For example, the screen data collection program generation unit 220 may index a main menu, a game start window, an in-game situation, and a game end situation.

The screen data collection program generating unit 220 sets a layer for each index after indexing of the video content is completed.

Specifically, the process of setting the layer for each index by the screen data collection program generating unit 220 is as follows.

First, the screen data collection program generating unit 220 sets a mask on the moving picture content, and sets the image resolution and the change characteristics for each resolution of each region in the mask in order to automatically catch the change of the region according to the change in resolution. Here, the screen data collection program generation unit 220 may select any one of linear and non-linear as a characteristic of change for each resolution of each region. In this case, the screen data collection program generating unit 220 may change the resolution through spline interpolation with respect to a resolution to which a coordinate value is not input.

In addition, the screen data collection program generation unit 220 selects a layer to be used for each index after setting the change characteristics for each resolution of each area. That is, the screen data collection program generating unit 220 selects a layer to be used by index in the layer selection window, and then drags it and moves it to the module creation tree. Here, the layer is basically a base layer applied just below the masking, cut a part of the region, detect the position values of the internal feature points, and compare the detected position values with the default value to mask the shaking of the region Image stabilization layer that compensates with data and image processing techniques, a translucent layer that exists to detect a translucent layer, and an area that operates independently of the previously used area, that is, an area with different change characteristics for each location and resolution. Additional layers may be included.

After the screen data collection program generation unit 220 completes the layer setting for each index, a portion having information to be recognized for each area is designated in a square or rectangular shape using a UI Location Filling Layer, and then The recognition module can be located by including it in a layer. As a result, the screen data collection program generation unit 220 can stably adjust the range to be recognized on the shake correction layer.

The screen data collection program generation unit 220 may set the type of information to be recognized through the recognition module after locating the region location filling layer.

After selecting the type of information to be recognized, the screen data collection program generating unit 220 inserts a recognition module into the area location filling layer. Here, the screen data collection program generation unit 220 may highlight and display a layer that can be used according to the type of previously input data, and may also automatically provide a recognition module corresponding to the area location filling layer. . For example, if the recognition range is determined through the area location filling layer, the recognition range may be in a square, circular, or rectangular shape. A recognition module can be automatically recommended and provided.

In addition, the screen data collection program generating unit 220 may provide help on the operation principle of each recognition module based on the characteristics of the layer.

In this way, the screen data collection program generation unit 220 may generate the screen data collection program through a series of procedures of indexing video content, setting layers for each index, and inserting a recognition module into each layer.

The storage unit 230 stores the image recognition program and at least one recognition module received from the web server 110 through the communication unit 210 , and the screen data collection program generated by the screen data collection program generation unit 220 . do. In addition, the storage unit 230 may store an operating system necessary for driving a screen data collection program and an image recognition program, and data required for an image recognition process and a screen data collection program generation process. For example, the storage 230 may store the above-described non-post-processed data and post-processed data. To this end, the storage unit 230 may be divided into a plurality of storage areas.

Also, the storage unit 230 may receive the updated recognition module from the web server 110 periodically or whenever the screen data collection program is executed.

The screen data collection unit 240 may recognize a large amount of data on a screen directly played by the user through the screen data collection program stored in the storage unit 230 .

Accordingly, according to the first embodiment of the present invention, the user can create a program for recognizing game data without complicated programming through the video content recognition device having the above configuration.

3 is a flowchart illustrating a video content recognition method according to the first embodiment of the present invention.

First, in order to recognize video content in the first embodiment of the present invention, three requirements including the current state and location of the region to be recognized, and information of the region to be recognized, that is, the data type of data, are essential. , the video content recognizing device 120 may place the recognition module in the area according to these three requirements.

In the embodiment of the present invention, for convenience of description, a game video is used as an example of video content, but the present invention is not limited thereto, and the video content may include various videos such as movies, sports, dramas, entertainment, and current affairs. In particular, in the embodiment of the present invention, the area of video content indicates a user interface (UI).

Referring to FIG. 3 , the apparatus 120 for recognizing video content according to the present invention accesses the web server 110 to recognize at least one recognition for recognizing information included in an image recognition program and video content from the web server 110 . A module is received (S310).

Next, the video content recognizing apparatus 120 designates the video content for which information is to be recognized on the operating system system through the received image recognition program (S320). In this case, the video content recognizing apparatus 120 cuts from the beginning of the video content to the end of which recognition is to be terminated by using an image editing tool.

Next, the video content recognizing apparatus 120 generates a screen data collection program by connecting at least one area for recognizing information in video content and at least one recognition module in the form of a layer (S330).

Finally, the video content recognizing apparatus 120 recognizes information included in the video content through the generated screen data collection program (S340).

4 is a flowchart illustrating a method of generating a screen data collection program according to the first embodiment of the present invention.

Referring to FIG. 4 , the video content recognizing apparatus 120 indexes video content in a time table ( S410 ). For example, the video content recognizing apparatus 120 may index a main menu, a game start window, an in-game situation, and a game end situation.

After indexing of the video content is completed, the video content recognizing apparatus 120 sets a layer for each index ( S420 ).

Next, the video content recognizing apparatus 120 recognizes by designating a part having information to be recognized for each area by using a UI Location Filling Layer after completing the layer setting for each index, and including it in an upper layer Position the module (S430). Here, the video content recognizing apparatus 120 may designate a part having information to be recognized in a square or rectangular shape.

The video content recognition apparatus 120 sets the type of information to be recognized through the recognition module after locating the area location filling layer (S440). For example, in-game data may include simple status information such as skill on/off or status abnormality, time-related data such as cooldown time or preparation time or respawn time, and quantitative data such as HP or item count. have.

Finally, after selecting the type of information to be recognized, the video content recognizing apparatus 120 inserts a recognition module into the area location filling layer ( S450 ). Here, the video content recognizing apparatus 120 may highlight and display a layer that can be used according to the type of previously input data, and may also automatically provide a recognition module corresponding to the area location filling layer. For example, if the recognition range is determined through the area location filling layer, the recognition range may be in a square, circular, or rectangular shape. can be automatically recommended and provided.

5 is a flowchart illustrating a method of setting a layer for each index according to the first embodiment of the present invention.

Referring to FIG. 5 , the moving image content recognizing apparatus 120 sets a mask on the moving image content ( S510 ).

The moving image content recognizing apparatus 120 sets the image resolution and the resolution characteristic of each region in the mask to automatically catch the change of the region according to the change in resolution (S520). Here, the moving image content recognizing apparatus 120 may select any one of linear and non-linear as a change characteristic for each resolution of each region. In this case, the video content recognizing apparatus 120 may change the resolution through spline interpolation for a resolution to which a coordinate value is not input.

The moving image content recognizing apparatus 120 selects a layer to be used for each index after setting the change characteristics for each resolution of each area (S530). That is, the video content recognizing apparatus 120 selects a layer to be used for each index in the layer selection window, and then drags it and moves it to the module creation tree. Here, the layer is basically the base layer applied immediately below the masking, cut a part of the region, detect the position values of the internal feature points, compare the detected position values with the default value, and then calculate the shaking of the region with the mask data and the image An additional layer that specifies an image stabilization layer that corrects with the processing technique, a translucent layer that exists for the detection of a translucent layer, and an area that operates independently of the previously used area, that is, an area with different change characteristics depending on location and resolution. may include

The above-described method may be implemented through various means. For example, embodiments of the present invention may be implemented by hardware, firmware, software, or a combination thereof.

In case of implementation by hardware, the method according to embodiments of the present invention may include one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), and Programmable Logic Devices (PLDs). , FPGAs (Field Programmable Gate Arrays), processors, controllers, microcontrollers and microprocessors, and the like.

In the case of implementation by firmware or software, the method according to the embodiments of the present invention may be implemented in the form of a module, procedure, or function that performs the functions or operations described above. The software code may be stored in the memory unit and driven by the processor. The memory unit may be located inside or outside the processor, and may transmit and receive data to and from the processor by various known means.

6 is a view showing an example screen of a time table according to the first embodiment of the present invention.

Referring to FIG. 6 , the user may index the main menu 610 , the game start window 620 , the in-game situation 630 , and the game end situation 630 in the time table. Here, the in-game situation 630 may include a survival situation and a death situation.

In addition, the setting of the time table varies in an area, and may be changed according to the overall setting of the recognition module, in-game play, game queue status, and the like.

7 is a view showing an example screen of the image recognition program according to the first embodiment of the present invention.

Referring to FIG. 7 , the user may drag each button 700 and bring it to the tree window 800 of the time table. In other words, if a lower layer is dragged onto an upper layer, it is included in that layer.

To this end, each layer has a priority, and in this case, there may be layers without priority. For example, the base layer 720 , the shake correction layer 730 , and the additional layer 740 have no priority, and the mask 710 , the base layer 720 , and the region location fill layer 750 have priority. There may be rankings. That is, the mask 710 may be an upper layer, the base layer 720 may be an intermediate layer, and the region location filling layer 750 may be a lower layer.

The base layer 720 serves to define regions including design elements (borders and boundaries, etc.) and information of the region, respectively, and classify each region according to the type of information that includes each region.

In the case of a game video, it is possible to set the area for each index by dividing the game situation under the time table, but nevertheless, there may be areas in which movement or color or graphic elements change under each other setting. For example, the skill window area and the minimap area are separated. If this is designated as a separate layer through the additional layer 740, values that change under different settings (eg, interface size adjustment and minimap size adjustment) may be applied respectively. That is, the additional layer 740 designates individual elements operating in a separate form in the video content.

As shown in FIG. 8 , the area location filling layer 750 serves to designate a pixel to be finally cropped in the capture area 810 of the upper layer, that is, an area to be input to the recognition module. That is, since the optimized pixel size for each open source is different in a square or rectangular shape, the user can designate an area that the recognition module can recognize through the area location filling layer 750 to fit each shape. Accordingly, the region location filling layer 750 may prevent different regions from encroaching on each other.

9 is a diagram showing a schematic configuration of the inside of a non-verbal information delivery device according to a second embodiment of the present invention.

The non-verbal information delivery device according to the present invention is worn on the wrist in the form of a strap, receives information recognized from the video content from the video content recognition device according to the first embodiment of the present invention, and tightens the received information; It is transmitted to the user in the form of pressure, electrical stimulation, and vibration.

9, the non-verbal information delivery device according to the present invention includes at least two or more solenoid modules 910, a pressure module 920, an electrical stimulation module 930, a vibration module 940, and a frequency control module ( 950 , a skin resistance measurement module 960 , and a controller 970 .

Each solenoid module 910 uses a neodymium magnet as a solenoid core, pushes the magnet by the repulsive force of the solenoid, and returns the magnet to its original position by the magnet's own magnetic force. At least two or more solenoid modules 910 having such a configuration may transmit information in a form of applying pressure to a moving direction or a rotating direction.

The pressure module 920 may be implemented as a linear servomotor, and operates on the principle of tightening the wrist by pulling a wire through the linear servomotor. For example, the pressure module 920 may gradually strengthen the tightening as the speed increases, and may gradually weaken the tightening as the speed decreases.

The electrical stimulation module 930 may be implemented as a module that generates a current of 1 to 2 mA in a high frequency form of 10 MHz or more. The frequency of the electrical stimulation module 930 may be adjusted through the resistance and capacitance values of the oscillation circuit.

The vibration module 940, like the solenoid module 910, is composed of a solenoid and a neodymium magnet, and operates by projecting a pulse to the solenoid to cause the neodymium magnet inside the solenoid to vibrate. The vibration intensity of the vibration module 940 may be adjusted by changing the frequency or adjusting the mass of the neodymium magnet by hardware or software.

The controller 970 may be implemented as a low-voltage, low-power MCU (Micro Control Unit) based on an Arduino. The control unit 970 supports 1 A battery charging, and the charging state can be checked through three LEDs. Also, the controller 970 may be equipped with a Bluetooth communication module and control at least 12 stimulation modules.

The control unit 970 according to the present invention having the above configuration determines whether the information of the moving image content received from the moving image content recognition apparatus is continuous information or single information, and if the recognized information is continuous information, the continuous information Information is transmitted in the form of pressure and tightening through the solenoid module 910 and the pressure module 920, respectively. In addition, when the recognized information is single information, the controller 970 transmits the single information in the form of vibration and electrical stimulation through each of the vibration module 940 and the electrical stimulation module 930 . To this end, the control unit 970 may allow the user to select a single piece of information through which module among the vibration module 940 and the electrical stimulation module 930 to transmit the single information.

On the other hand, in the non-verbal information delivery device according to the present invention, even if the same amount of current flows according to the frequency, the degree of percutaneous, dermal stimulation and muscle stimulation varies, so a frequency control module 950 that can adjust this is further added. may include The frequency control module 950 adjusts the frequency by adjusting the capacitance value of the oscillation circuit in the electrical stimulation module 930 .

In addition, the non-verbal information delivery device according to the present invention may further include a skin resistance measurement module 960 for measuring the resistance of the skin in order to flow a constant current because the skin resistance is not always constant. . Accordingly, the control unit 970 may adjust the frequency of the electrical stimulation module 930 through the frequency adjustment module 950 according to the resistance value measured through the skin resistance measurement module 960 .

10 is a view showing the configuration of a solenoid module according to a second embodiment of the present invention.

Referring to FIG. 10 , the solenoid module 910 according to the present invention includes a solenoid 912 and a magnet 914 . The magnet 914 is a neodymium magnet and is used as a core of the solenoid 912 .

As shown in (b) of FIG. 10, the solenoid module 910 pushes the magnet 914 by the repulsive force of the solenoid 912, and as shown in FIG. The magnet 914 is returned to its original position by its own magnetic force. At least two solenoid modules 910 having such a configuration may transmit information in the form of applying pressure to the moving direction or the rotating direction.

Referring to FIG. 11 , the non-verbal information delivery device according to the present invention receives information recognized from video content from the video content recognizing device 120 ( S1110 ).

The non-verbal information delivery device determines whether the received information is continuous information or single information (S1120). For example, when the video content is a game video, a skill state including a stun may be single information, and a skill or a buff's cool time may be continuous information.

When the received information is continuous information, the non-verbal information transmitting apparatus transmits the continuous information in the form of pressure and tightening through the solenoid module 910 and the pressure module 920, respectively (S1130).

When the received information is single information, the non-verbal information delivery device transmits the single information in the form of vibration and electrical stimulation through the vibration module 940 and the electrical stimulation module 930, respectively (S1140).

On the other hand, although not shown in FIG. 11 of the present invention, the non-verbal information transmission method according to the present invention includes the steps of: the non-verbal information transmission device measuring the resistance of the skin through the skin resistance measurement module 960; , adjusting the frequency of the electrical stimulation module 930 through the frequency adjustment module 950 according to the resistance value measured through the skin resistance measurement module 960 may be further included.

The embodiments disclosed herein have been described above with reference to the accompanying drawings. As such, the embodiments shown in each drawing should not be construed as being limited, and may be combined with each other by those skilled in the art having read the contents of the present specification, and when combined, it may be construed that some components may be omitted.

Here, the terms or words used in the present specification and claims should not be construed as being limited to conventional or dictionary meanings, but should be interpreted as meanings and concepts consistent with the technical ideas disclosed in the present specification.

Therefore, the embodiments described in the present specification and the configurations shown in the drawings are only the embodiments disclosed in the present specification, and do not represent all the technical ideas disclosed in the present specification, so various equivalents that can replace them at the time of the present application It should be understood that there may be water and variations.

The present invention can be used to easily tune and produce a program for recognizing information included in video content by a plurality of users through a graphical user interface environment.

Claims

a web server for storing a plurality of recognition modules for recognizing information included in video content; and

Receive at least one recognition module from among the plurality of recognition modules from the web server, designate the video content for which the information is to be recognized, and form at least one region of the video content and the at least one recognition module in the form of a layer a video content recognition device for recognizing information included in the video content through the at least one recognition module connected to

An image processing system comprising a.
According to claim 1,

a plurality of cross-validation devices for inputting various types of information included in the video content into a recognition module and performing cross-validation on the result values of the recognition module to update the plurality of recognition modules;

Image processing system, characterized in that it further comprises.
a storage unit for storing at least one recognition module for recognizing information included in video content; and

Designate video content for which information is to be recognized, connect at least one region of the video content and the at least one recognition module stored in the storage in a layer form, and include in the video content through the at least one recognition module a control unit for recognizing the specified information;

A video content recognition device comprising a.
4. The method of claim 3,

The control unit indexes the video content in a time table, sets layers for each index, and inserts the recognition module into each layer to generate the screen data collection program.
5. The method of claim 4,

and the control unit sets a mask on the video content, sets an image resolution and a change characteristic for each resolution of each region in the mask, and then selects a layer to be used for each index.
6. The method of claim 5,

The control unit selects any one of linear and non-linear as a change characteristic for each resolution of each region.
7. The method of claim 6,

Wherein the control unit changes the resolution based on the coordinate value input from the user when non-linear is selected as the change characteristic for each resolution of each region.
8. The method of claim 7,

The controller changes the resolution through spline interpolation with respect to a resolution to which a coordinate value is not input.
6. The method of claim 5,

The layers to be used for each index include a base layer, a shake correction layer that cuts off a part of an area, detects the position values of internal feature points, and compares the detected position values with the default value to correct the shake of the area, a translucent layer, and a position , a moving image content recognition apparatus comprising at least one of additional layers for designating regions having different change characteristics for each resolution.
6. The method of claim 5,

The controller designates an area to which the recognition module is input on the layer through a UI Location Filling Layer, and inserts the recognition module into the area location filling layer. .
In the video content recognition method of a video content recognition device for recognizing information included in video content,

receiving and storing at least one recognition module for recognizing information included in the video content;

designating the video content for which the information is to be recognized; and

recognizing information included in the video content through the at least one recognition module by connecting at least one region of the video content and the at least one recognition module in a layered form;

A method for recognizing video content, including
The method of claim 11, wherein recognizing the information included in the video content comprises:

indexing the video content in a time table;

setting a layer by index; and

inserting the recognition module into each layer;

Video content recognition method comprising a.
The method of claim 12, wherein the setting of the layer by index comprises:

setting a mask on the video content;

setting an image resolution and a change characteristic for each resolution of each region in the mask; and

selecting a layer to use by index;

Video content recognition method comprising a.
The method of claim 13, wherein the setting of the change characteristics for each resolution comprises:

selecting any one of linear and non-linear as a change characteristic for each resolution of each region; and

changing a resolution based on a coordinate value input from a user when non-linearity is selected as a change characteristic for each resolution of each region;

Video content recognition method comprising a.
The method of claim 12, wherein the inserting of the recognition module comprises:

designating an area into which the recognition module is input on the layer through a UI Location Filling Layer; and

inserting the recognition module into the region location filling layer;

Video content recognition method comprising a.