KR20240053888A - Filtering technology and method for blocking viewing of harmful videos in user device - Google Patents

Abstract

The present invention applies when a user receives a video file from the Internet or an external storage medium to a user terminal, watches a video through streaming, or watches a video file stored inside the user terminal or on an external storage device connected to the user terminal. By extracting the unique characteristics of the video, determining whether the video is illegal or harmful, blocking or disabling it, and storing and disclosing the information and inflow path on a remote public DB server, the video can be distributed socially or In order to block the inflow path of the relevant video and thereby fundamentally block social and economic damage caused by illegal and harmful videos, priority is given to online or user terminals 100 such as the Internet or streaming; an image inflow unit (10) through which images flow into the user terminal from an internal or external storage device connected to the user terminal; a feature extraction module unit (20) for extracting unique features of the image; a feature comparison module unit 200 that determines whether the extracted features are illegal or harmful; a harmful substance characteristic information DB unit 300 that stores in advance characteristic information of illegal or harmful substances that can be compared by the characteristic comparison module unit from a third input means; An inflow viewing blocking module unit (30) that stops the inflow or viewing of illegal or harmful content; An information collection module unit (40) that collects information about the video and the path through which the video flows when the inflow or viewing is blocked; It consists of a blocking information public DB unit (400) that stores the collected information to block illegal or harmful videos from entering or viewing from the user terminal, and the information is stored in the DB and disclosed to the public to fundamentally prevent social and economic damage. It is done.

Description

{Filtering technology and method for blocking viewing of harmful videos in user device}

The present invention extracts the unique characteristics of the video when a user receives a video file from the Internet or an external storage medium, watches a video through streaming, or watches a video file stored in a user terminal. By determining whether it is illegal or harmful, blocking or disabling it, and storing and disclosing the information and inflow path on a remote public server, the video is distributed socially or the inflow path of the video is blocked, thereby preventing social damage caused by illegal and harmful videos. It concerns technologies and methods to prevent economic damage at the source.

Since the Internet and wireless communication became active, illegal video recordings such as hidden cameras and revenge videos and their illegal distribution have been flooding in, and the social damage caused by illegal videos, such as the N-room incident, is spreading day by day. To solve and prevent this problem, the government Various laws have been enacted, management and technical measures are mandatory for related businesses, and efforts are being made to provide relief to victims. Public institutions and civic groups are constantly monitoring online to detect related illegal videos and harmful materials.

However, in the existing method of blocking illegal harmful videos, the website or app that distributes the video is found through manual search or victim report, and the relevant Internet address (URL) is registered in a public database to block access to the network. The method is not only difficult to respond immediately despite the mobilization of numerous organizations and personnel and a lot of effort and cost, but it is also virtually useless, as it can no longer be blocked when the URL is changed or is ineffective on overseas sites. Video feature-based filtering technology (video DNA filtering), which is a technical measure of the N-room prevention law proposed as an alternative, is a form in which website and app servers provide the function of extracting and comparing video DNA. This is done by accessing the website or app. It operates in a form of blocking only when videos are downloaded or uploaded, failing to solve the problem of having to manually register the video DNA information database of illegal harmful videos that are compared, using a lot of manpower and cost, and the website. If the app does not use video DNA filtering technology, there is a problem that it does not work at all, so there is also a famous problem with overseas sites or P2P file sharing services or chat apps such as Telegram that do not provide it. As a result, even after the N-room incident, secondary and tertiary social problems such as the L-room incident continue to occur and the number of victims increases every year, and the social problems caused by illegal harmful videos are not resolved.

The present invention is intended to solve the above problems. Among the video DNA filtering technologies, the video DNA extraction module is operated in real time in conjunction with the video input unit in the user terminal, so that the video inflow is stopped or viewing is blocked after it is determined whether the video is illegally harmful. By taking actions such as these immediately at the point when the video is imported or played on the user terminal, downloading and uploading of illegal harmful videos as well as streaming and playback can be blocked at the user terminal level, and P2P file sharing and P2P file sharing as well as websites and apps. It will be possible to fundamentally block the distribution of illegal harmful videos online, such as blocking distribution on chat apps such as Telegram at the user terminal level.

In addition, in this process, information that can only be extracted from the terminal, such as the relevant video information and the inflow path to the terminal, is transmitted in real time to the network blocking information public DB, and the accumulated information in the public DB is analyzed in real time to distribute the URL information that distributes the target video. By automatically updating in real time to a public DB, it is possible to provide a network blocking effect that can respond in real time to circumvention measures such as changes to overseas sites and URLs without large costs and manpower consumption.

As another purpose of the present invention, when linking the video characteristic (DNA) information DB of copyright-protected items such as movies and dramas with the feature comparison module instead of the harmful substance characteristic information DB, the copyright of digital content that is indiscriminately illegally copied and distributed online Infringements can also be prevented at the stage of user terminal use. In addition, not only for the purpose of blocking illegal harmful images or digital copyright infringement as described above, but also for criminal investigation and criminal investigation and copyright infringement information DB and copyright infringement information DB collected in the process are provided to government agencies such as prosecutors and police. Even in prevention, value can be created as evidence.

The present invention primarily includes an image inflow unit where images flow into the user terminal from online, such as the Internet or streaming, or from an external storage device connected to the user terminal, a feature extraction module unit for extracting unique features of the image, and an extracted The feature comparison module unit determines whether the feature is illegal or harmful, and the feature comparison module unit stores the characteristic information of the illegal or harmful material in advance from a third input means, which can be compared to the harmful material characteristic information DB section, which is imported if the feature is illegal or harmful. An inflow viewing blocking module unit that stops video or viewing, an information collection module unit that collects information about the video and the path through which the video flows when inflow or viewing is blocked, and a blocking information public that stores the collected information. This is achieved by configuring the DB part.

As another aspect of the present invention, the video begins to be copied, transmitted, downloaded, and streamed online, such as the Internet, and begins to be stored in the memory of the user terminal, or the video is played from inside the user terminal or from an external storage medium connected to the user terminal. A video inflow stage in which a video file is opened in the user terminal when an event such as an upload attempt, copy, or harmfulness test begins; A feature extraction step of extracting unique features (image DNA) of the image; A feature comparison step of comparing and searching the extracted image DNA with a remote harmful substance characteristic information database to determine whether the image corresponds to an illegal harmful substance; At this time, if it is an illegal harmful material, a process interruption step to stop the video inflow such as transmission, download, streaming, or playback; A path extraction step of extracting the path through which the video came in, such as the external URL or IP address from which the video is downloaded or streamed, the name of the web or app, or the name of the chat room; A progress path storage step of storing the extracted progress route and the information of the corresponding video in a remote public blocking information DB; A hazard level analysis step for each path that analyzes the stored path and classifies the hazard level; an analysis result storage step of storing the analyzed results; An inflow continuation step of continuing the inflow if the image introduced in the hazardous material determination step is not a harmful material; This is achieved by configuring a storage or playback stage that completes the purpose of inflow: transmission or downloading, streaming, playback, etc.

As described above, the present invention improves the feature-based (DNA) video filtering technology to fundamentally block illegal harmful videos on user terminals and blocks the distribution of illegal harmful videos by updating the distribution route information of illegal harmful videos to the public database in real time. It is a technology that can protect users.

1 is a configuration diagram showing the relationship between each function of the filtering technology and method for blocking viewing of harmful videos in a user terminal of the present invention;
Figure 2 is a flowchart showing the operation method of the harmful video viewing blocking filtering technology and method in the user terminal of the present invention;
Figure 3 shows the operation method of the harmful video viewing blocking filtering technology and method in the user terminal of the present invention when video is imported from the Internet or external storage media.
flowchart showing,
Figure 4 is a flowchart showing the operation method of the filtering technology and method for blocking harmful video viewing in a user terminal of the present invention when playing video from an internal storage device in the user terminal or when a similar event occurs;
Figure 5 is a configuration diagram showing the relationship between each function for digital copyright protection of the filtering technology and method for blocking harmful video viewing in the user terminal of the present invention.

Embodiments of the present invention will be described in detail with reference to the accompanying drawings as follows.

1 is a configuration diagram showing the relationship between each function of the filtering technology and method for blocking viewing of harmful videos in a user terminal of the present invention;

Figure 2 is a flowchart showing the operation method of the harmful video viewing blocking filtering technology and method in the user terminal of the present invention;

Figure 3 is a configuration diagram showing the relationship between each function for digital copyright protection of the harmful video viewing blocking filtering technology and method in the user terminal of the present invention

Figure 1 is a configuration diagram showing the relationship between each function of the harmful video viewing blocking filtering technology and method in the user terminal of the present invention. The present invention provides a user terminal 100, such as online or streaming, such as the Internet or streaming; an image inflow unit (10) through which images flow into the user terminal from an internal or external storage device connected to the user terminal; a feature extraction module unit (20) for extracting unique features of the image; a feature comparison module unit 200 that determines whether the extracted features are illegal or harmful; a harmful substance characteristic information DB unit 300 that stores in advance characteristic information of illegal or harmful substances that can be compared by the characteristic comparison module unit from a third input means; An inflow viewing blocking module unit (30) that stops the inflow or viewing of illegal or harmful content; An information collection module unit (40) that collects information about the video and the path through which the video flows when the inflow or viewing is blocked; It consists of a blocking information public DB unit (400) that stores the collected information to block illegal or harmful videos from entering or viewing from the user terminal, and the information is stored in the DB and disclosed to the public to fundamentally prevent social and economic damage. It is a technical feature that does this.

As shown in FIG. 1, the image inflow unit 10, the feature extraction module 20, the inflow viewing blocking module 30, and the information collection module 40 operate inside the user terminal 100 and collect harmful substance characteristic information. DB (300) is linked from an external remote location. The feature comparison module 200 may be linked from an external remote location or may operate inside the user terminal 100 as needed. The Harmful Material Characteristics Information DB (300) refers to an information DB accumulated by third-party information inputters such as victims, civic groups, and public institutions in the government or public institutions. For example, broadcasting under the Korea Communications Commission This may include the public database of information on the characteristics of illegally filmed videos operated by the review committee.

In addition, when a user wants to download or watch a video online or through an external storage device, when the video begins to flow into the video inlet 10 inside the user terminal 100, the feature extraction module 20 detects the video. Unique features (image DNA) are extracted and the image DNA is transmitted to the feature comparison module 200, and the feature comparison module 200 searches and compares the hazardous substance characteristic information DB 300 to determine whether the image DNA is registered as an illegal harmful substance. judge. If the video DNA is judged to be illegally harmful, the inflow viewing blocking module 30 stops the download or playback in progress in the video inlet 10, and the information collection module 40 collects the type, file name, and hash of the video. Information such as capacity, video DNA, etc., and which website or app the video was transmitted or streamed from. Internet address (URL) or IP, site or app name, web page name, web page publisher, chat room name, creator, P2P name. , hash files, log files, etc. are collected. The inflow path collection information is transmitted to a remote blocking information public DB (400) and accumulated, and the blocking information public DB (400) is based on all data such as the frequency, harmfulness level, and damage information of the inflow path information. The hazards of the route are analyzed, classified, and stored. The blocking information public DB (400) is disclosed to the public in real time and is used to recognize the harmfulness of websites or apps in the inflow path in the network and terminal of an unspecified third party and block the network in real time. Additionally, it is possible to raise awareness by displaying a message warning the user that the current video has been determined to be illegal and harmful and will be sent to the public blocking information database.

Figure 2 is a flow chart showing the operating method of the harmful video viewing blocking filtering technology and method in the user terminal of the present invention. The present invention begins with copying, transmitting, downloading, and streaming videos from online, such as the Internet, and stores them in the memory of the user terminal. A video import step in which the video file is opened in the user terminal when storage begins, the video is played from inside the user terminal or an external storage medium connected to the user terminal, or an event such as an upload attempt or harmfulness test begins; A feature extraction step of extracting unique features (image DNA) of the image; A feature comparison step of comparing and searching the extracted image DNA with a remote harmful substance characteristic information database to determine whether the image corresponds to an illegal harmful substance; At this time, if it is an illegal harmful material, a process interruption step to stop the video inflow such as transmission, download, streaming, or playback; A path extraction step of extracting the path through which the video came in, such as the external URL or IP address from which the video is downloaded or streamed, the name of the web or app, or the name of the chat room; A progress path storage step of storing the extracted progress route and the information of the corresponding video in a remote public blocking information DB; A hazard level analysis step for each path that analyzes the stored path and classifies the hazard level; an analysis result storage step of storing the analyzed results; An inflow continuation step of continuing the inflow if the image introduced in the hazardous material determination step is not a harmful material; It consists of a storage or playback stage that completes the purpose of inflow, such as transmission or downloading, streaming, and playback.

In addition, in the feature comparison stage, where the video DNA extracted through the feature extraction step is searched and compared with a remote harmful material characteristic information database to determine whether the video corresponds to illegal harmful material, if the video is not illegal harmful material, the inflow of the video is prevented. It consists of a continuous stage where it is stored or played on the user terminal.

In addition, as shown in Figure 2, the video import step is specifically a process of starting to record the video data in the memory of the user terminal by copying, transmitting, downloading, streaming, etc. online such as the Internet, and the user terminal It consists of the process of opening the video file to play the video internally or through an external storage device connected to the user terminal, or to proceed with events such as upload attempts, copying, and harmfulness testing.

Figure 3 is a configuration diagram showing the relationship between each function for digital copyright protection of the harmful video viewing blocking filtering technology and method in the user terminal of the present invention. The present invention includes an online user terminal 100 such as the Internet or streaming; an image inflow unit (10) through which images flow into the user terminal from an internal or external storage device connected to the user terminal; a feature extraction module unit (20) for extracting unique features of the image; a feature comparison module unit 200 that determines whether the extracted features are illegal or harmful; a copyright-protected content DB unit 300 that stores in advance feature information of copyright-protected content that can be compared by the feature comparison module unit from a third input means; An inflow viewing blocking module unit 30 that stops the inflow or viewing of content that violates copyright protection; An information collection module unit (40) that collects information about the video and the path through which the video flows when the inflow or viewing is blocked; It consists of a copyright infringement information DB unit 400 that stores collected information, blocking the inflow or viewing of copyright infringement videos from user terminals, and fundamentally preventing economic damage to copyright holders by storing the information in the DB and disclosing it to relevant organizations. Its technical feature is to protect digital copyrights.

As shown in FIG. 3, the image inflow unit 10, the feature extraction module 20, the inflow viewing blocking module 30, and the information collection module 40 operate inside the user terminal 100 and store copyright-protected content. DB (300) is linked from an external remote location. The feature comparison module 200 may be linked from an external remote location or may operate inside the user terminal 100 as needed. Copyright protection content DB (300) refers to an information DB that accumulates information about digital content subject to copyright protection by the copyright holder or a person delegated by the copyright holder. For example, the copyright database operated by the Korea Copyright Commission may correspond to this. There may be a third party.

In addition, when a user wants to download or watch a video online or through an external storage device, when the video begins to flow into the video inlet 10 inside the user terminal 100, the feature extraction module 20 detects the video. Unique features (image DNA) are extracted and the image DNA is transmitted to the feature comparison module 200, and the feature comparison module 200 searches and compares the copyright protection content DB 300 to determine whether the image DNA infringes copyright. judge. If the video DNA is determined to be a copyright violation video, the inflow viewing blocking module 30 stops the download or playback in progress in the video inlet 10, and the information collection module 40 determines the type and file name of the video, Information such as hash, capacity, video DNA, and which website or app the video was transmitted or streamed from. Internet address (URL) or IP, site or app name, web page name, web page publisher, chat room name, creator, P2P Collect the inflow path of the video, including names, hash files, and log files. The inflow path collection information is transmitted to the remote copyright infringement information DB (400) and accumulated, and the copyright infringement information DB (400) is based on all data such as frequency, harmfulness, and damage information of the inflow path information. The illegality of the route is analyzed, classified, and stored. The copyright infringement information DB (400) is disclosed to relevant organizations, such as the copyright holder or a person delegated by the copyright holder, so that it can be used for the copyright holder's copyright protection activities. Additionally, it is possible to raise awareness by printing a message warning users that the current video has been judged to be a copyright violation video and that the copyright holder will be notified.

Hereinafter, the operation and effects of the present invention will be described in detail with reference to FIGS. 1 to 3.

The above configuration (applicant by the present applicant?), Republic of Korea Patent Publication No. 1015370320000 (publication date July 15, 2015) is a method of generating image feature information for image filtering and image features based on the recording medium on which the method is recorded. As an improved information (image DNA) filtering technology, the present invention, unlike existing technologies and methods, operates the image DNA extraction module in the image DNA filtering technology in real time in conjunction with the image inlet in the user terminal to prevent illegal harmful effects of the image. After determining whether or not video is being imported, measures such as stopping the video inflow or blocking viewing are taken immediately at the point when the video is imported or played on the user terminal, so that not only downloading and uploading of illegal harmful videos, but also streaming and playback can be blocked at the user terminal level. , it has the ability to fundamentally block the distribution of illegal harmful videos online, such as blocking distribution not only on websites and apps, but also on P2P file sharing and chatting apps such as Telegram, at the user terminal level.

In addition, in this process, information that can only be extracted from the terminal, such as the relevant video information and the inflow path to the terminal, is transmitted in real time to the network blocking information public DB, and the accumulated information in the public DB is analyzed in real time to distribute the URL information that distributes the target video. By automatically updating in real time to a public DB, it is possible to provide a network blocking effect that can respond in real time to circumvention measures such as changes to overseas sites and URLs without large costs and manpower consumption.

In fact, the effect of the present invention is explained by comparing the examples of existing technology and the present invention as follows.

Previously, there was a harmful substance blocking technology that collects information on sites designated as harmful, such as pornographic sites and gambling sites, provides it to a public database, and protects users by blocking Internet addresses (URLs) from the network and preventing users from accessing those addresses. In this process, a lot of manpower and money are spent by governments and private organizations to collect information on harmful sites. However, when the harmful site is blocked, the operator of the harmful site can immediately change to a different Internet address and continue to operate, while the changed Internet address is also updated in the public database only after a lot of manpower and time are spent. In addition, countless harmful sites appear from time to time around the world, and it was not enough to manually monitor them with manpower and cost. Therefore, the reality is that existing harmful content blocking technology is ineffective, and anyone can freely access harmful information with a quick search.

In addition, existing harmful content blocking technology had no way to prevent the distribution of harmful videos even if it blocked pornographic sites. Even if a specific site is blocked, harmful videos that have already been distributed can be exchanged anywhere online, such as general community sites, P2P, KakaoTalk, and chat apps such as Telegram. As a result, illegal footage has recently been distributed through Telegram, causing a large number of There was no way to prevent the N-room incident, which created a large number of victims and became a social problem.

To complement this, a method for generating image feature information for image filtering of the recent Korean Patent Publication No. 1015370320000 (published on July 15, 2015) and an image feature information (video DNA) filtering technology based on the recording medium on which the method is recorded are developed. was invented, and a similar type of technology was adopted, and the N-room Prevention Act was enacted and technical measures were made mandatory. However, this is a method of providing video DNA filtering technology from the service provider side, that is, the server side, not the user terminal, and the technology is provided to the server. If not applied, the distribution of illegal harmful videos is still possible through all other servers, and there is a limitation that harmful information in public databases still has to be manually registered by human resources, making real-time response difficult. In other words, even if the government or agency monitors and discovers illegally filmed content, the damage will continue to spread while the information is updated in the public database and applied to filtering, and if filtering technology is not applied to all online services, it will still be a useless method. It is.

The present invention completely improves the above problems by operating video DNA filtering technology in the user terminal and fundamentally blocking the inflow of illegal harmful videos when an attempt is made, and updating the public database with the inflow path of the video in real time. This is a way to do it. In other words, when a user attempts to download an illegal harmful image from an Internet site called A and a chat room B, and the image flows into the user's terminal and begins to be stored in the memory, the feature extraction module of the present invention extracts the DNA of the image and makes it public. It determines in real time whether the video is illegal and harmful from the DB, and if it is illegal, it blocks the video from flowing into the user's terminal and transmits information from site A and chat room B, which are the inflow channels of the video, to the public DB in real time. As a result, users are fundamentally unable to access videos, and the public database is automatically updated in real time by each user across the country. In addition to blocking illegal harmful videos, information on websites, apps, and chat rooms that provide them are also provided in a public database, making it possible to block networks in real time for harmful sites, and furthermore, can be used as criminal evidence for the distribution of illegal videos.

10: Image inlet 11:
12:
20: Feature extraction module 21:
30: Incoming viewing blocking module 31:
40: Information collection module 41:
100: user terminal
200: Feature comparison module
300: Harmful substance characteristic information DB
400: Block information public DB

Claims (7)

A main server linked to an advertising DB where advertisements and advertisement information are accumulated;
Each is connected to or included in one or more Internet sites that provide services using the Internet, and the product DB and member DB of the Internet site are linked and connected to the main server, and the advertisement and advertisement information are sent to the main server. It consists of a client server that mediates the Internet site and the main server and constitutes an advertising site by transmitting customer information and product information at the same time,
When a user accesses the Internet site and receives a service or purchases a product, an advertising list optimized for the user is linked from the main server and the client server, so that the user can select an advertisement instead of watching the advertisement according to the user's will. An advertising and discount system using the Internet, characterized by selling at a discounted price. It connects an existing or new Internet site, a main server that is linked to an advertising database with accumulated advertisements and advertising information, and a client server that mediates the Internet site and the main server and constitutes an advertising site, and connects member information, product information, A first step of linking advertisements and advertisement information data with each other;
When a user accesses the Internet site and selects a product to purchase a product and requests a discount, the client server analyzes the user attribute information of the user and the product attribute information of the product and transmits it to the main server. Step 2 and;
A third step of creating a list of advertisements applicable to the user and the product in the advertisement DB of the main server based on the data transmitted by the client server, and transmitting the information list of the advertisements to the client server;
a fourth step in which the client server outputs an advertisement according to a customer's discount request based on the advertisement information list transmitted from the main server and discounts the price of the product according to the advertisement information;
In the fourth step above, when the minimum price of the product is reached or a customer requests a purchase, the price of the product is determined at that price, a payment function page is printed, and after payment, a database related to sales and customers is created. step;
Advertising and discount method using the Internet, characterized in that it consists of: The method of claim 2, wherein the first step is
A DB linking step of linking the member DB and product DB of each site in an existing or new Internet site to the client server, and transmitting the corresponding information from the client server to the main server;
An advertisement input step in which the operator of the main server inputs the ordered advertisement and information about the advertisement into the advertisement DB of the main server;
An information integration step of building a DB by integrating the information in the main server based on the product and member information transmitted from the client and the input advertisement information;
Advertising and discount method using the Internet, characterized in that it consists of. The method of claim 3, wherein the information integration step is
calculating the total number of advertisements and sales for each period, and the price and number of products eligible for discount to derive the lowest possible discount amount for each product;
placing an appropriate advertisement for each product according to the properties of the input advertisement and the properties of the product;
Classifying the level of each client server according to its contribution to the overall system, such as the number of members and connections of the client server, and determining a discount application level for the product accordingly;
Advertising and discount method using the Internet, characterized in that it consists of. The method of claim 2, wherein the fourth step is
Step 4-1 of outputting an optimal advertisement to the user when the user requests a discount;
Step 4-2 of discounting the product purchase price according to the number and time of advertisements viewed by the user;
Step 4-3 of storing advertisements and discount information viewed by the user;
Afterwards, step 4-4 of repeating the above step or proceeding to a lower step depending on whether there is a remaining advertisement;
Advertising and discount method using the Internet, characterized in that it consists of. According to claim 5, between steps 4-1 and 4-2, the user does not comply with the mandatory viewing time when watching the advertisement, or does not respond to the questionnaire in the case of a survey advertisement, etc. Advertising using the Internet, characterized in that step 4-1a is additionally configured to repeat step 4-1 without proceeding with step 4-2 if the advertisement viewing requirements included in the advertising information are not met, and Discount method. The method of claim 2, wherein in the fourth step, when an advertisement is output to the user and the user watches the advertisement, the discount rate of the product selected by the user is graphically indicated using a price bar that graphically indicates the discount rate of the product. Advertising and discount method using the Internet, characterized by: