WO2020093757A1

WO2020093757A1 - Risk detection method, apparatus and system based on background collaboration

Info

Publication number: WO2020093757A1
Application number: PCT/CN2019/102129
Authority: WO
Inventors: 李东声
Original assignee: 天地融科技股份有限公司
Priority date: 2018-11-06
Filing date: 2019-08-23
Publication date: 2020-05-14
Also published as: CN111144181A

Abstract

The present invention provides a risk detection method, apparatus and system based on the background collaboration. The method comprises: a first camera sends first video data to a monitoring apparatus; a second camera sends second video data to the monitoring apparatus, the first camera and the second camera being provided at different positions in the environment to be detected; the monitoring apparatus recognizes the face corresponding to a user to be analyzed in the first video data and the second video data to determine said user, obtains video data comprising said user at the required point in the first video data and the second video data, extracts the background feature from the video data comprising said user at the required point, inputs the background feature obtained by extracting into a preset background collaboration model, and calculates to obtain the matching degree between the background feature and the preset background collaboration model; the monitoring apparatus compares the matching degree with a preset background threshold, if the matching degree is less than the preset background threshold, determines that the preset risk exists.

Description

Risk detection method, device and system based on background collaboration

Cross-reference of related applications

This application requires the priority of the Chinese patent application number "201811311971.2" filed by Tiandirong Technology Co., Ltd. on November 6, 2018, with the invention titled "a risk detection method, device and system based on background collaboration".

Technical field

The invention relates to the field of video surveillance, and in particular to a method, device and system for risk detection based on background collaboration.

Background technique

The existing Automatic Teller Machine (ATM) is generally installed in a self-service bank. After inserting a bank card on the ATM, you can perform bank counter services such as withdrawals, deposits, and transfers on the ATM. Due to the openness, convenience and environmental particularity of self-service banks and ATMs. In recent years, criminal activities against self-service banks and ATMs have been increasing.

However, the traditional ATM video surveillance system mainly records the video. After the incident, the recorded video is forensic afterwards, which can eliminate disputes and crack cases. However, such a mechanism can only provide the effect of forensic afterwards, and cannot do it. Real-time or early warning.

Summary of the invention

The present invention aims to solve the above problem / one.

The main purpose of the present invention is to provide a risk detection method, device and system based on background collaboration.

To achieve the above objective, the technical solution of the present invention is specifically implemented as follows:

One aspect of the present invention provides a risk detection method based on background collaboration, including: a first camera performs video collection on a detection environment, obtains first video data, and sends the first video data to a monitoring device; a second camera detects The environment collects video, obtains second video data, and sends the second video data to the monitoring device, where the first camera and the second camera are set at different locations in the environment to be detected; the monitoring device receives the first video data and the first Two video data, identify the face corresponding to the user to be analyzed in the first video data and the second video data, and determine the user to be analyzed; the monitoring device obtains the first video data and the second video data contains the user to be analyzed is located at the mandatory point Video data at the location, and extract background features from the video data containing the user at the mandatory point to be analyzed; the monitoring device inputs the extracted background features into the preset background collaboration model, and calculates the background features and the preset background collaboration The matching degree between the models; the monitoring device will match the matching degree with the preset background threshold Comparison of line, if the matching degree is lower than a preset threshold background, generating a first comparison result, determines the default risk exists.

Another aspect of the present invention provides a risk detection system based on background collaboration, including: a first camera, used for video collection of a detection environment, obtaining first video data, and sending the first video data to a monitoring device; Two cameras, used for video collection of the environment to be detected, obtaining second video data, and sending the second video data to the monitoring device, wherein the first camera and the second camera are set at different positions in the environment to be detected; the monitoring device For receiving the first video data and the second video data, identifying the face corresponding to the user to be analyzed in the first video data and the second video data, and determining the user to be analyzed; acquiring the first video data and the second video data Contains the video data of the user to be analyzed at the mandatory point, and extracts background features from the video data of the user to be analyzed at the mandatory point; input the extracted background feature into a preset background collaborative model to calculate the background feature The matching degree with the preset background collaborative model; the matching degree is performed with the preset background threshold More, if the matching degree is lower than a preset threshold background, generating a first comparison result, determines the default risk exists.

Another aspect of the present invention provides a risk detection device based on background collaboration, including: a receiving module, configured to receive a first camera to perform video acquisition on the environment to be detected and obtain first video data; and receive a second camera to perform the environment to be detected Video capture, obtain the second video data, and send the second video data to the monitoring device, wherein the first camera and the second camera are set at different positions in the environment to be detected; the determination module is used to identify the first video The face corresponding to the user to be analyzed in the data and the second video data determines the user to be analyzed; the extraction module is used to obtain the video data of the first video data and the second video data containing the user to be analyzed at the mandatory point, and Extract background features from the video data containing the user at the mandatory point to be analyzed; a calculation module for inputting the extracted background features into a preset background collaborative model, between the calculated background features and the preset background collaborative model Matching degree; judgment module for comparing the matching degree with the preset background threshold, such as BACKGROUND matching degree is lower than a preset threshold value, generating a first comparison result, determines the default risk exists.

It can be seen from this that the risk detection method, device and system based on background collaboration provided by the embodiments of the present invention set at least two cameras at different positions to identify people, and analyze the background characteristics of the user when passing the required point Through analysis, it is possible to discover preset risks in real time (such as illegal and criminal intent), and solve the drawbacks of deliberate counterfeiting and other criminal behaviors that were unavoidable under the supervision of separate cameras in the past.

BRIEF DESCRIPTION

In order to more clearly explain the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. Those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts.

1 is a flowchart of a risk detection method based on background collaboration provided by an embodiment of the present invention;

2 is a schematic structural diagram of a risk detection system based on background collaboration provided by an embodiment of the present invention;

FIG. 3 is a schematic structural diagram of a risk detection device based on background collaboration provided by an embodiment of the present invention.

detailed description

The embodiments of the present invention will be further described in detail below with reference to the drawings.

FIG. 1 shows a flowchart of a risk detection method based on background collaboration provided by an embodiment of the present invention. Referring to FIG. 1, a risk detection method based on background collaboration provided by an embodiment of the present invention includes:

S101: The first camera performs video acquisition on the environment to be detected, obtains first video data, and sends the first video data to the monitoring device; the second camera performs video acquisition on the environment to be detected, obtains second video data, and transmits the second video The data is sent to the monitoring device, wherein the first camera and the second camera are set at different positions in the environment to be detected.

Specifically, the first camera and the second camera are cameras set at different positions in the environment to be detected. For example, when the environment to be detected is a self-service bank, the first camera may be a camera set on the ATM machine, and the second camera may be An environmental camera installed in an environment other than an ATM in a self-service bank. Of course, in the actual application of the present invention, more than two cameras may be provided, which is not limited in the present invention.

The first camera and the second camera capture the video at the mandatory point from different locations and have different background characteristics. Among them, the mandatory point is the point that the user must pass when entering the environment to be processed in the to-be-detected environment. In the embodiment of the present invention, the mandatory point can be set in advance, and at the same time, the mandatory point can be one or more. No specific limitation is made in the present invention. It is worth noting that, because the positions of the first camera and the second camera are different, the same user may only be captured by one of the first camera or the second camera when passing the same mandatory point.

The first video data collected by the first camera and the second video data collected by the second camera are sent to the monitoring device in real time, or the collected video data are regularly sent to the monitoring device according to a preset period.

S102. The monitoring device receives the first video data and the second video data, recognizes the face corresponding to the user to be analyzed in the first video data and the second video data, and determines the user to be analyzed.

Specifically, the monitoring device may be installed near the camera or in the background. For example, in a self-service banking environment, it can be installed in the ATM machine or in the bank monitoring background, which is not specifically limited in the present invention. After receiving the first video data and the second video data, the monitoring device uses face recognition technology to identify a user from the first video data and a user from the second video data, and determines that the two users are the same As a user, it is determined that the user is a user to be analyzed.

If the monitoring device only recognizes one user in the first video data or the second video data, the user may be directly regarded as the user to be analyzed.

If the user identified by the monitoring device from the first video data and the user identified by the second video data are not the same user, the two users are treated as different users to be analyzed for analysis.

S103. The monitoring device acquires the video data including the user to be analyzed at the mandatory point in the first video data and the second video data, and extracts background features from the video data including the user at the mandatory point.

Specifically, because the video data of the user at the mandatory point often contains the user to be analyzed, which is valid data, the video data at the non-mandatory point may not contain the user to be analyzed, which is invalid data, and the invalid data is analyzed. It does not make much sense for risk detection. Therefore, the monitoring device performs background feature extraction only for the video data where the user is at the mandatory point, so as to avoid extracting background features in invalid data and improve the efficiency of risk detection.

The background feature may include any feature of the background marker in the environment and any combination thereof to play a role of identifying the background. For example, it may contain information on the position of static objects, shape information of static objects, information on the number of static objects, and the movement laws of dynamic objects.

S104: The monitoring device inputs the extracted background features to a preset background collaborative model, and calculates the matching degree between the background features and the preset background collaborative model.

Specifically, the background collaboration model is preset in the monitoring device to analyze the background features. As an optional implementation of the embodiment of the present invention, the monitoring device receives training video data collected by the first camera and the second camera in advance; the monitoring device extracts training elements from the training video data separately, and obtains presets based on the training elements. Background collaborative model. By analyzing the background markers in the shooting range of each camera, a background collaborative model is generated, and a reasonable background threshold range is set according to the mandatory points in different movement trajectories of normal users to improve the intelligence and accuracy of the judgment.

In specific applications, the ATM camera and the environmental camera transmit the video captured in the monitoring range to the monitoring device. The monitoring system extracts the background markers 1, 2 ... n, analyzes and calculates the background collaborative model, and the user reaches the ATM from different paths , Through different cameras, the monitoring device analyzes the background markers extracted from each path, sets a reasonable background threshold and a reasonable background judgment method, so as to establish the preset background collaborative model of the present invention, and set A reasonable background threshold and a reasonable background determination method can be set accordingly according to different application scenarios, which is not specifically described in the present invention.

The extracted background features are input into a preset background collaborative model, and the matching degree between the extracted background features and the background collaborative model is calculated. The matching degree is a numerical value, for example, it can be a percentage value.

S105. The monitoring device compares the matching degree with a preset background threshold. If the matching degree is lower than the preset background threshold, a first comparison result is generated to determine that there is a preset risk.

Specifically, when the matching degree is lower than the preset background threshold, it is considered that the background feature does not match the background collaborative model, and if the background feature does not match the background collaborative model, the preset risk is considered to exist, for example: extraction The video with this background feature is at risk or the user to be analyzed is at risk, for example, the video is tampered with, the camera is hijacked, or the user disrupts the normal collection of the camera. As an optional embodiment of the present invention, when the matching degree is not lower than a preset background threshold, the monitoring device generates a second comparison result to determine that there is no preset risk. Since the matching degree between the background features and the background collaborative model is high enough, it can be considered that there is no risk, for example, there is no risk in the video or there is no risk in the user to be analyzed.

In specific applications, for example, in a self-service banking environment, when a user arrives at the ATM, the monitoring device performs background analysis based on the received video containing user characteristics, inputs the background characteristics of the user in each camera into the background collaborative model, and matches the output The degree is compared with the background threshold to obtain a comparison result 1, so as to determine whether there is a risk according to the comparison result 1.

Optionally, as an optional embodiment of the present invention, the monitoring device performs an alarm operation after determining that the user to be analyzed has a preset risk. The alarm operation can be an alarm device in the environment to be detected, for example, by sounding a light-emitting alarm, or an alarm device in the monitoring room of the background monitoring personnel, for example, by displaying an alarm or audible alarm on the monitoring display screen, or sending a short message Alarm to the monitoring personnel or police personnel. By alerting when a risk occurs, the efficiency of the risk management of self-service banks and ATMs is further improved.

It can be seen that through the risk detection method based on background collaboration provided by the embodiment of the present invention, at least two cameras are set at different positions to identify persons, and by analyzing the background characteristics of the user when passing the required point to be analyzed, it is possible to Discover preset risks in real time (such as illegal and criminal intent), and solve the shortcomings of deliberate forgery and other crimes that were unavoidable under the supervision of separate cameras in the past.

As an optional embodiment of the present invention, the first video data collected by the first camera is encrypted by a security chip provided in the first camera, and the second video data collected by the second camera is provided by the second camera The security chip is encrypted, the first camera sends the encrypted first video data to the monitoring device, and the second camera sends the encrypted second video data to the monitoring device; the monitoring device receives the encrypted first video data and encryption After the second video data, decrypt the encrypted first video data and the encrypted second video data to obtain the first video data and the second video data. Encrypted transmission of video data improves the security of video data transmission and prevents video data from being tampered with after being cracked.

The first video data collected by the first camera is signed by a security chip provided in the first camera to obtain first signature data, and the second video data collected by the second camera is signed by a security chip provided in the second camera Obtain the second signature data, the first camera sends the first video data and the first signature data to the monitoring device, the second camera sends the second video data and the second signature data to the monitoring device; the monitoring device receives the first video data After verifying the first signature data and the second video data and the second signature data, verify the first signature data and the second signature data, and use the first video data and the second video data for subsequent analysis after the signature is passed . By signing video data, you can ensure the authenticity of the source of the video data and prevent the video data from being tampered with.

FIG. 2 shows a schematic structural diagram of a risk detection system based on background collaboration provided by an embodiment of the present invention. The risk detection system based on background collaboration provided by an embodiment of the present invention applies the above-mentioned method. The structure of the risk detection system is briefly described. For other unfinished matters, refer to the above description of the risk detection method based on background collaboration. Referring to FIG. 2, the risk detection system based on background collaboration provided by an embodiment of the present invention includes:

The first camera 201 is used to collect video for the environment to be detected, obtain first video data, and send the first video data to the monitoring device;

The second camera 202 is used to collect video in the environment to be detected, obtain second video data, and send the second video data to the monitoring device, wherein the first camera and the second camera are set at different positions in the environment to be detected;

The monitoring device 203 is used to receive the first video data and the second video data, identify the face corresponding to the user to be analyzed in the first video data and the second video data, determine the user to be analyzed; obtain the first video data and the second The video data includes the video data of the user to be analyzed at the mandatory point, and extracts the background features from the video data of the user at the mandatory point; input the extracted background features into the preset background collaborative model and calculate Obtain the matching degree between the background feature and the preset background collaborative model; compare the matching degree with the preset background threshold, and if the matching degree is lower than the preset background threshold, generate a first comparison result and determine that there is a preset risk.

It can be seen that, through the risk detection system based on background collaboration provided by the embodiment of the present invention, at least two cameras are set at different positions to identify persons, and by analyzing the background characteristics of the user when passing the mandatory point to be analyzed, it can be Discover preset risks in real time (such as illegal and criminal intent), and solve the shortcomings of deliberate forgery and other crimes that were unavoidable under the supervision of separate cameras in the past.

As an optional embodiment of the present invention, the monitoring device 203 is further configured to generate a second comparison result when the matching degree is not lower than a preset background threshold to determine that there is no preset risk. Since the matching degree between the background features and the background collaborative model is high enough, it can be considered that there is no risk, for example, there is no risk in the video or there is no risk in the user to be analyzed.

As an optional embodiment of the present invention, the monitoring device 203 is further used to receive training video data collected by the first camera and the second camera in advance; extract training elements from the training video data separately, and obtain presets based on the training elements. Background collaborative model. By analyzing the background markers in the shooting range of each camera, a background collaborative model is generated, and a reasonable background threshold range is set according to the mandatory points in different movement trajectories of normal users to improve the intelligence and accuracy of the judgment.

As an optional embodiment of the present invention, it is characterized in that the monitoring device 203 is also used to perform an alarm operation after determining that the user to be analyzed has a preset risk. By alerting when a risk occurs, the efficiency of the risk management of self-service banks and ATMs is further improved.

As an optional embodiment of the present invention, the first video data collected by the first camera 201 is encrypted by a security chip provided in the first camera, and the second video data collected by the second camera 202 is provided by the second The security chip in the camera is encrypted, the first camera 201 sends the encrypted first video data to the monitoring device, and the second camera 202 sends the encrypted second video data to the monitoring device 203; the monitoring device 203 receives the encrypted After the first video data and the encrypted second video data, the encrypted first video data and the encrypted second video data are decrypted to obtain the first video data and the second video data. Encrypted transmission of video data improves the security of video data transmission and prevents video data from being tampered with after being cracked.

The first video data collected by the first camera 201 is signed by a security chip provided in the first camera to obtain the first signature data, and the second video data collected by the second camera 202 is passed through the security chip provided in the second camera Signing to obtain the second signature data, the first camera 201 sends the first video data and the first signature data to the monitoring device, the second camera 202 sends the second video data and the second signature data to the monitoring device 203; the monitoring device 203 After receiving the first video data and the first signature data and the second video data and the second signature data, verify the first signature data and the second signature data, and use the first video data and the second signature data after passing the verification Second video data for subsequent analysis. By signing the video data, you can ensure the authenticity of the source of the video data and prevent the video data from being tampered with.

On the basis of FIG. 2, FIG. 3 shows a schematic structural diagram of a risk detection device based on background collaboration provided by an embodiment of the present invention. The risk detection device based on background collaboration is a monitoring device in the system shown in FIG. 2. The risk detection device based on background collaboration provided by the embodiment of the invention applies the above system and method. The following only briefly describes the structure of the risk detection device based on background collaboration provided by the embodiment of the present invention. For a related description of the risk detection system and method, referring to FIG. 3, a risk detection device based on background collaboration provided by an embodiment of the present invention includes:

The receiving module 2031 is configured to receive the first video data acquired by the first camera for video acquisition of the environment to be detected; receive the second camera acquire the second video data for video acquisition of the environment to be detected, and send the second video data to Monitoring device, wherein the first camera and the second camera are set at different positions in the environment to be detected;

The determining module 2032 is configured to identify the face corresponding to the user to be analyzed in the first video data and the second video data, and determine the user to be analyzed;

An extraction module 2033 is used to obtain video data containing the user to be analyzed at the mandatory point in the first video data and the second video data, and extract background features from the video data containing the user at the mandatory point to be analyzed;

The calculation module 2034 is configured to input the extracted background features into a preset background collaborative model, and calculate the matching degree between the background features and the preset background collaborative model;

The judgment module 2035 is configured to compare the matching degree with a preset background threshold, and if the matching degree is lower than the preset background threshold, generate a first comparison result and determine that there is a preset risk.

It can be seen that through the risk detection device based on background collaboration provided by the embodiment of the present invention, at least two cameras are set at different positions to identify people, and by analyzing the background characteristics of the user when passing the required point to be analyzed, it is possible to Discover preset risks in real time (such as illegal and criminal intent), and solve the shortcomings of deliberate forgery and other crimes that were unavoidable under the supervision of separate cameras in the past.

As an optional implementation manner of the present invention, the judgment module 2035 is further configured to generate a second comparison result when the matching degree is not lower than a preset background threshold to determine that there is no preset risk. Since the matching degree between the background features and the background collaborative model is high enough, it can be considered that there is no risk, for example, there is no risk in the video or there is no risk in the user to be analyzed.

Although the embodiments of the present invention have been shown and described above, it can be understood that the above-mentioned embodiments are exemplary and cannot be understood as limitations to the present invention, and those of ordinary skill in the art will not deviate from the principles and purposes of the present invention The above embodiments can be changed, modified, replaced and modified within the scope of the present invention. The scope of the invention is defined by the appended claims and their equivalents.

Claims

A risk detection method based on background collaboration is characterized by including:

The first camera collects video from the environment to be detected, obtains the first video data, and sends the first video data to the monitoring device;

The second camera performs video acquisition on the environment to be detected, obtains second video data, and sends the second video data to the monitoring device, wherein the first camera and the second camera are disposed at the to-be-detected Different locations in the environment;

The monitoring device receives the first video data and the second video data, recognizes a face corresponding to the user to be analyzed in the first video data and the second video data, and determines the user to be analyzed;

The monitoring device acquires the video data including the user to be analyzed at the mandatory point in the first video data and the second video data, and Extract background features from video data;

The monitoring device inputs the extracted background features into a preset background collaboration model, and calculates the matching degree between the background features and the preset background collaboration model;

The monitoring device compares the matching degree with a preset background threshold, and if the matching degree is lower than the preset background threshold, generates a first comparison result and determines that there is a preset risk.
The method according to claim 1, further comprising:

When the matching degree is not lower than the preset background threshold, the monitoring device generates a second comparison result to determine that there is no preset risk.
The method according to claim 1 or 2, further comprising:

The monitoring device receives training video data collected by the first camera and the second camera in advance;

The monitoring device separately extracts training elements from the training video data, and trains to obtain the preset background collaborative model according to the training elements.
The method according to claim 1 or 2, further comprising:

After determining that the user to be analyzed has a preset risk, the monitoring device performs an alarm operation.
A risk detection system based on background collaboration is characterized by including:

The first camera is used to collect video in the environment to be detected, obtain first video data, and send the first video data to a monitoring device;

The second camera is used to collect video in the environment to be detected, obtain second video data, and send the second video data to the monitoring device, wherein the first camera and the second camera are provided in the Describe the different locations in the environment to be tested;

The monitoring device is configured to receive the first video data and the second video data, identify the face corresponding to the user to be analyzed in the first video data and the second video data, and determine the Analyzing users; acquiring the first video data and the second video data including the video data of the user to be analyzed at the mandatory point, and from the video including the user to be analyzed at the mandatory point Extract background features from the data; input the extracted background features into a preset background collaborative model, and calculate the matching degree between the background features and the preset background collaborative model; The set background threshold is compared, and if the matching degree is lower than the preset background threshold, a first comparison result is generated to determine that there is a preset risk.
The system according to claim 5, wherein the monitoring device is further configured to generate a second comparison result to determine that there is no preset risk when the matching degree is not lower than the preset background threshold .
The system according to claim 5 or 6, wherein the monitoring device is further configured to receive training video data collected by the first camera and the second camera in advance; from the training video data Extract training elements separately, and train to obtain the preset background collaborative model according to the training elements.
The system according to claim 5 or 6, wherein the monitoring device is further configured to perform an alarm operation after determining that the user to be analyzed has a preset risk.
A risk detection device based on background collaboration is characterized in that it includes:

The receiving module is configured to receive the first video data obtained by the first camera for video collection of the environment to be detected; receive the second camera to collect the second video data for video collection of the detected environment, and send the second video data To the monitoring device, wherein the first camera and the second camera are provided at different positions in the environment to be detected;

A determining module, configured to identify a face corresponding to the user to be analyzed in the first video data and the second video data, and determine the user to be analyzed;

An extraction module for acquiring video data including the user to be analyzed at the mandatory point in the first video data and the second video data, and from the user at the mandatory point including the to-be-analyzed user Extract background features from the video data;

A calculation module, configured to input the extracted background feature into a preset background collaborative model, and calculate a matching degree between the background feature and the preset background collaborative model;

The judgment module is configured to compare the matching degree with a preset background threshold, and if the matching degree is lower than the preset background threshold, generate a first comparison result and determine that there is a preset risk.
The apparatus according to claim 1, wherein the judgment module is further configured to generate a second comparison result when the matching degree is not lower than the preset background threshold to determine that there is no preset risk .