Real-time assessment method for leakage risk of visible privacy information of ground target party caused by suspicious unmanned aerial vehicle

Abstract

The invention provides a real-time assessment method for a risk of leakage of visible privacy information of a ground target party caused by a suspicious unmanned aerial vehicle. The method comprises the following steps: S1, evaluating the leakage criticality level of visible privacy information of a ground target party; s2, recording the starting time of the generation process of the visible privacy information of the ground target as a time point 0, and determining a time interval T2 of real-time evaluation; then the step S3 and the step S4 are operated in sequence every T2; s3, at the time point t, performing fth real-time evaluation on the leakage probability level of the visible privacy information of the ground target party; and S4, performing fth real-time evaluation on the leakage risk level of the visible privacy information of the ground target party at the time point t, and stopping evaluation until the generation process of the visible privacy information of the ground target party is finished. According to the method, the research blank of the risk assessment problem of leakage of visible privacy information of the ground target party caused by the suspicious unmanned aerial vehicle is filled, and the assessment method has good rationality and feasibility.

Description

Technical Field

[0001] The present invention relates to the technical field of drone security risk assessment, and in particular to a real-time assessment method for the risk of privacy information leakage visible to ground targets caused by suspicious drones. Background Art

[0002] Drone accidents can be divided into active accidents intentionally caused by drone operators and passive accidents unintentionally caused by operators. The two types of accidents cause active safety risks and passive safety risks respectively. The parties involved in active drone accidents can be divided into destroyers and targets, and the parties involved in passive drone accidents can be divided into flight operators, flight customers, and third parties that have no direct interest in the flight. According to the consequences of drone accidents, drone safety risks can be divided into casualty risks, property loss risks, and privacy information leakage risks. At present, there are drone safety risk assessment methods for ground objects. These methods include: ground third-party casualty and property loss risk assessment methods, ground target casualty and property loss risk assessment methods, however, there is currently no relevant solution to assess the risk of leakage of visible privacy information of ground targets, that is, the assessment of the risk of leakage of visible privacy information of ground targets is completely a research gap.

[0003] In view of this research gap, it is extremely necessary to design a real-time assessment method for the risk of privacy information leakage visible to ground targets caused by suspicious drones. Visible privacy information refers to privacy information that can be collected by devices that form and record images based on the principle of optical imaging. Summary of the invention

[0004] The present invention aims to propose a real-time assessment method for the risk of private information leakage visible to ground targets caused by suspicious drones. The assessment method is implemented by a risk assessment party. When a suspicious drone appears in the airspace around a ground object, the risk assessment party assumes that the suspicious drone may cause an active accident, and regards the suspicious drone as the destroyer and the ground object as the ground target party, and conducts a real-time assessment of the risk of private information leakage visible to the ground target party.

[0005] To this end, the purpose of the present invention is to propose a real-time assessment method for the risk of privacy information leakage visible to ground targets caused by suspicious drones.

[0006] In order to achieve the above-mentioned purpose, the technical solution of the present invention provides a method for real-time assessment of the risk of private information leakage visible to ground targets caused by suspicious drones. The method for real-time assessment of the risk of private information leakage visible to ground targets caused by suspicious drones includes: step S1: assessing the level of hazard of leakage of private information visible to ground targets; step S2: recording the start time of the process of generating private information visible to ground targets as time point 0, and determining the time interval T of real-time assessment 2 ; Then every T 2 , that is, at time point t, step S3 and step S4 are executed in sequence; t = (f-1) × T 2 ; f represents the number of real-time evaluations, and f is a positive integer greater than or equal to 1; step S3: at time point t, perform the f-th real-time evaluation on the leakage probability level of the privacy information visible to the ground target party; step S4: based on the evaluation result of the leakage hazard level of the privacy information visible to the ground target party and the f-th real-time evaluation result of the leakage probability level of the privacy information visible to the ground target party, perform the f-th real-time evaluation on the leakage risk level of the privacy information visible to the ground target party at time point t, and stop the evaluation until the generation process of the privacy information visible to the ground target party is completed.

[0007] Preferably, the step S1 comprises: step S1.1: constructing a network model of the organization where the relevant persons of the ground target party are located; wherein the relevant persons of the ground target party refer to persons associated with the ground target party; the relevant persons of the ground target party are confirmed by the risk assessment party; step S1.2: calculating the importance of each member in the organization where the relevant persons of the ground target party are located based on the network model; step S1.3: calculating the ranking ratio value of the importance of the relevant persons of the ground target party based on the calculation result of the importance of each member in the organization where the relevant persons of the ground target party are located; step S1.4: determining the leakage hazard level of the visible privacy information of the ground target party based on the calculation result of the ranking ratio value of the importance of the relevant persons of the ground target party.

[0008] Preferably, the step S3 comprises: step S3.1: at time point t, monitoring the surrounding airspace of the ground target party, and executing steps S3.2 to S3.4 according to the monitoring results, or skipping steps S3.2 and S3.3 and directly executing step S3.4; step S3.2: at time point t, calculating the probability of the suspicious drone identifying the identity of the ground target party; step S3.3: at time point t, calculating the clarity of the privacy information visible to the ground target party photographed by the suspicious drone; step S3.4: at time point t, calculating the leakage probability of the privacy information visible to the ground target party; step S3.5: at time point t, performing the fth real-time evaluation of the leakage probability level of the privacy information visible to the ground target party.

[0009] Preferably, the step S1.1 specifically includes: step S1.11: the risk assessment party determines the organization of the person related to the ground target party according to the daily social behavior of the person related to the ground target party; constructs the network model of the organization of the person related to the ground target party as a two-way complex network; wherein each member in the organization of the person related to the ground target party is modeled as a two-way complex network node; step S1.12: statistics the information interaction between internal members of the organization of the person related to the ground target party within a certain time span; if member i sends information to member j, then establish an edge from node i to node j, and use the number of sent information as the weight W of the edge. ij ; Step S1.13: Establish a super node, establish edges from the super node to other nodes in the bidirectional complex network, and set the weights of the edges to 1; Step S1.14: Set the initial value of the importance S value of each node in the bidirectional complex network to 1;

[0010] The step S1.2 specifically includes:

[0011] Step S1.21: Iteratively calculate the importance S value of each node. When the importance S value of each node converges, the iterative calculation ends. The formula corresponding to the iterative calculation is:

[0012]

[0013] In formula (1), S j (tt) represents the importance of node j calculated at the (tt)th iteration; S i (tt+1) represents the importance of node i calculated at the (tt+1)th iteration; the S value corresponding to each member node is the importance of each member in the organization where the person related to the ground target party is located; m represents the total number of edges with node i as the end point in the network model; W ji W represents the weight of the edge with starting point j and end point i; jp represents the weight of the edge with starting point j and end point p; q represents the total number of edges with node j as the starting point in the network model; p represents the end point of the edge with node j as the starting point in the network model;

[0014] The step S1.3 specifically includes: step S1.31: according to the calculation result of step S1.21, the importance of each member in the organization of the ground target party related person is arranged in descending order; step S1.32: the ranking number of the ground target party related person is recorded as l, the total number of organizational personnel is recorded as L, and the ranking ratio value of the importance of the ground target party related person is calculated; the formula corresponding to the calculation of the ranking ratio value is:

[0015]

[0016] And the step S1.4 specifically includes: step S1.41: according to the calculation result z of step S1.32, determine the leakage hazard level of the privacy information visible to the ground target party; if z∈[0,0.06), the leakage hazard level of the privacy information visible to the ground target party is "serious"; if z∈[0.06,0.125), the leakage hazard level of the privacy information visible to the ground target party is "relatively serious"; if z∈[0.125,0.25), the leakage hazard level of the privacy information visible to the ground target party is "medium"; if z∈[0.25,0.5), the leakage hazard level of the privacy information visible to the ground target party is "relatively minor"; if z∈[0.5,1], the leakage hazard level of the privacy information visible to the ground target party is "minor".

[0017] Preferably, the step S3.1 specifically includes: using detection equipment to monitor whether there is a suspicious drone in the surrounding airspace of the ground target; if the monitoring result shows that there is a suspicious drone, then setting p t =1, and record the location of the suspicious drone; if the monitoring result shows that there is no suspicious drone, then set p t =0; if p t =1, then continue to execute steps S3.2 to S3.4; if p t = 0, then let q t =0, and skip steps S3.2 and S3.3, and directly execute step S3.4;

[0018] The step S3.2 specifically includes: Step S3.21: Determine the shooting angle of the nth suspicious drone on the ground target identification marker information surface in, is the horizontal shooting angle, is the vertical shooting angle, and the ground target party identity identification marker information surface is confirmed by the risk assessment party; Step S3.22: by measuring the brightness δ of the ground target party identity identification marker information surface t , the occlusion ratio γ of the ground target identification marker information surface t , environmental visibility level κ t ; where κ t ∈[0,4],γ t ∈[0,1]; Step S3.23: According to the type of ground target identification marker, select and use the pre-trained ground target identification probability evaluation neural network model to evaluate the ground target identification probability p obtained by the nth suspicious drone t n ; Step S3.24: At time point t, the formula corresponding to the probability of the suspicious drone identifying the ground target is:

[0019]

[0020] In formula (3), p t represents the probability of the suspicious drone identifying the ground target at time t; represents the probability of ground target identification obtained by the nth suspicious UAV at time τ; τ represents the evaluation time point identifier, and its value range is all evaluation time points from time 0 to time t; τ∈{0,T 2 ,T 2 *2,T 2 *3...,t};

[0021] Step S3.25: If the ground target does not carry an identification marker, but the risk assessment party infers the probability p of the suspicious drone obtaining the current ground target's identity through existing flight activities based on the existing evidence * , then we can t Direct assignment, that is, p t =p * ;

[0022] The step S3.3 specifically includes: Step S3.31: Determine the shooting angle of the nth suspicious drone on the ground target information plane in, Represents the horizontal shooting angle, represents the vertical shooting angle; the ground target information plane refers to the plane presenting the privacy information of the ground target, which is determined by the risk assessment party; step S3.32: determining the brightness of the ground target by measuring t , ground target information surface occlusion ratio Ambient visibility level κ t ; where κ t ∈[0,4], Step S3.33: According to the type of ground target, select and use the pre-trained ground target visible privacy information clarity assessment neural network model to assess the clarity of the privacy information visible to the ground target photographed by the nth suspicious drone Step S3.34: At time point t, the formula corresponding to the clarity of the private information visible to the ground target photographed by the suspicious drone is obtained as follows:

[0023]

[0024] The step S3.4 specifically includes: Step S3.41: The risk assessment party estimates the estimated length T of the process of generating the visible privacy information of the ground target party according to the real-time situation 1 ; Step S3.42: Based on the estimated length T1 And formula (5) to calculate the leakage probability of private information visible to the ground target;

[0025]

[0026] In formula (5), is the estimated number of monitoring time points included in the process of generating visible privacy information; T 2 represents the time interval of real-time evaluation; P t represents the probability of leakage of private information visible to the ground target at time t; mm is a variable representing an integer, and its value range is from 0 to t / T 2 Integer: mm∈{0,1,...,t / T 2};

[0027] And the step S3.5 specifically includes: step S3.51: according to the calculation result of the step S3.42, at time point t, the leakage probability level of the visible privacy information of the ground target is evaluated for the fth time in real time; if P t ∈[0,0.1), then the fth real-time evaluation result of the probability level of leakage of the visible privacy information of the ground target is “slight”; if P t ∈[0.1,0.4), then the ground target can see that the fth real-time evaluation result of the probability level of leakage of privacy information is “relatively slight”; if P t ∈[0.4,0.6), then the fth real-time evaluation result of the leakage probability level of the visible privacy information of the ground target is “medium”; if P t ∈[0.6,0.9), the ground target can see that the fth real-time evaluation result of the probability level of privacy information leakage is “more serious”; if P t ∈[0.9,1], the f-th real-time evaluation result of the probability level of leakage of visible privacy information to the ground target is “serious”.

[0028] Preferably, according to the evaluation result of the leakage hazard level of the privacy information visible to the ground target party obtained in step S1.41 and the f-th real-time evaluation result of the leakage probability level of the privacy information visible to the ground target party obtained in step S3.51, and according to the risk matrix shown in Table 1, the leakage risk level of the privacy information visible to the ground target party is evaluated for the f-th time in real time at time point t;

[0029] Table 1 Risk Matrix

[0030]

[0031] Beneficial effects of the present invention:

[0032] (1) The real-time assessment method for the risk of private information leakage visible to ground targets caused by suspicious drones provided by the present invention provides a specific and feasible solution to the research gap of the risk assessment problem of private information leakage visible to ground targets caused by suspicious drones. The proposed solution follows the risk assessment framework and conducts real-time assessment of the risk level according to the probability of visible private information leakage and the degree of harm of visible private information leakage, which has sufficient theoretical basis.

[0033] (2) The real-time assessment method for the risk of leakage of visible privacy information of the ground target party caused by a suspicious drone provided by the present invention aims to solve the problem that the hazard level of visible privacy information leakage is difficult to measure. The importance ranking ratio of the ground target party’s related persons is calculated based on a complex network node importance evaluation algorithm, and is used as the basis for assessing the hazard level of leakage of visible privacy information of the ground target party. The assessment method has good rationality and feasibility.

[0034] (3) The present invention provides a real-time assessment method for the risk of private information leakage visible to the target party on the ground caused by a suspicious drone. According to the process of the private information leakage accident visible to the target party on the ground, a probability assessment model for the private information leakage visible to the target party on the ground is established. The assessment model has good rationality.

[0035] Additional aspects and advantages of the invention will become apparent from the following description, or may be learned by practice of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A schematic flow chart of a method for real-time assessment of privacy information leakage risk visible to ground targets caused by a suspicious drone according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0037] In order to more clearly understand the above-mentioned objectives, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0038] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0039] Figure 1 The following is a schematic flow chart showing a method for real-time assessment of the risk of privacy information leakage visible to ground targets caused by suspicious drones according to an embodiment of the present invention. Figure 1 As shown in the figure, the real-time assessment method of the risk of privacy information leakage caused by the suspicious drone to the ground target is as follows:

[0040] S1: Evaluate the level of hazard of leakage of private information visible to ground targets;

[0041] S2: Determine time point 0 and real-time evaluation time interval T 2 ; The first evaluation is performed at t=0, and then every T 2 , that is, at time point t, S3 and S4 are run in sequence;

[0042] S3: Step S3: At time point t, the leakage probability level of the private information visible to the ground target party is evaluated in real time;

[0043] Step S4: at time point t, the leakage risk level of the private information visible to the ground target party is evaluated in real time until the generation process of the private information visible to the ground target party is completed and the evaluation is stopped.

[0044] In this embodiment, the start time of the privacy information generation process is 0, the first evaluation is performed at t=0, and then it is performed every T2 until the privacy information generation process is completed and the evaluation is stopped.

[0045] Specifically, in one embodiment of the present invention, the real-time assessment method for the risk of leakage of private information visible to the ground target party caused by the suspicious drone includes: step S1: assessing the leakage risk level of the private information visible to the ground target party; step S2: recording the start time of the process of generating the private information visible to the ground target party as time point 0, and determining the time interval T of the real-time assessment 2 ; Then every T 2 , that is, at time point t, step S3 and step S4 are executed in sequence; t = (f-1) × T 2 ; f represents the number of real-time evaluations, and f is a positive integer greater than or equal to 1; step S3: at time point t, perform the f-th real-time evaluation on the leakage probability level of the privacy information visible to the ground target party; step S4: based on the evaluation result of the leakage hazard level of the privacy information visible to the ground target party and the f-th real-time evaluation result of the leakage probability level of the privacy information visible to the ground target party, perform the f-th real-time evaluation on the leakage risk level of the privacy information visible to the ground target party at time point t, and stop the evaluation until the generation process of the privacy information visible to the ground target party is completed.

[0046] In this embodiment, the real-time assessment method for the risk of privacy information leakage visible to ground targets caused by suspicious drones provided by the present invention provides a specific and feasible solution to the research gap in the risk assessment problem of privacy information leakage visible to ground targets caused by suspicious drones. The proposed solution follows the risk assessment framework and conducts real-time assessment of the risk level according to the probability of visible privacy information leakage and the degree of hazard of visible privacy information leakage, and has sufficient theoretical basis.

[0047] In one embodiment of the present invention, the step S1 includes: step S1.1: constructing a network model of the organization where the ground target party-related person is located; wherein the ground target party-related person refers to a person associated with the ground target party; the ground target party-related person is confirmed by the risk assessment party; step S1.2: based on the network model, calculating the importance of each member in the organization where the ground target party-related person is located; step S1.3: based on the calculation result of the importance of each member in the organization where the ground target party-related person is located, calculating the ranking ratio value of the importance of the ground target party-related person; step S1.4: based on the calculation result of the ranking ratio value of the importance of the ground target party-related person, determining the leakage hazard level of the visible privacy information of the ground target party.

[0048] In one embodiment of the present invention, the step S3 includes: step S3.1: at time point t, monitor the surrounding airspace of the ground target party, and execute steps S3.2 to S3.4 according to the monitoring results, or skip steps S3.2 and S3.3 and directly execute step S3.4; step S3.2: at time point t, calculate the probability of the suspicious drone identifying the identity of the ground target party; step S3.3: at time point t, calculate the clarity of the visible privacy information of the ground target party photographed by the suspicious drone; step S3.4: at time point t, calculate the leakage probability of the visible privacy information of the ground target party; step S3.5: at time point t, perform the fth real-time evaluation of the leakage probability level of the visible privacy information of the ground target party.

[0049] In one embodiment of the present invention, the step S1.1 specifically includes: step S1.11: the risk assessment party determines the organization of the person related to the ground target party based on the daily social behavior of the person related to the ground target party; constructs a network model of the organization of the person related to the ground target party as a two-way complex network; wherein each member of the organization of the person related to the ground target party is modeled as a two-way complex network node; step S1.12: statistics on the information interaction between internal members of the organization of the person related to the ground target party within a certain time span; if member i sends information to member j, an edge from node i to node j is established, and the number of sent information is used as the weight W of the edge. ij ; Step S1.13: Establish a super node, establish edges from the super node to other nodes in the bidirectional complex network, and set the weights of the edges to 1; Step S1.14: Set the initial value of the importance S value of each node in the bidirectional complex network to 1;

[0050] The step S1.2 specifically includes:

[0051] Step S1.21: Iteratively calculate the importance S value of each node. When the importance S value of each node converges, the iterative calculation ends. The formula corresponding to the iterative calculation is:

[0052]

[0053] In formula (1), S j (tt) represents the importance of node j calculated at the (tt)th iteration; S i (tt+1) represents the importance of node i calculated at the (tt+1)th iteration; the S value corresponding to each member node is the importance of each member in the organization where the person related to the ground target party is located; m represents the total number of edges with node i as the end point in the network model; W ji W represents the weight of the edge with starting point j and end point i; jp represents the weight of the edge with starting point j and end point p; q represents the total number of edges with node j as the starting point in the network model; p represents the end point of the edge with node j as the starting point in the network model;

[0054] The step S1.3 specifically includes: step S1.31: according to the calculation result of step S1.21, the importance of each member in the organization of the ground target party related person is arranged in descending order; step S1.32: the ranking number of the ground target party related person is recorded as l, the total number of organizational personnel is recorded as L, and the ranking ratio value of the importance of the ground target party related person is calculated; the formula corresponding to the calculation of the ranking ratio value is:

[0055]

[0056] And the step S1.4 specifically includes: step S1.41: according to the calculation result z of step S1.32, determine the leakage hazard level of the privacy information visible to the ground target party; if z∈[0,0.06), the leakage hazard level of the privacy information visible to the ground target party is "serious"; if z∈[0.06,0.125), the leakage hazard level of the privacy information visible to the ground target party is "relatively serious"; if z∈[0.125,0.25), the leakage hazard level of the privacy information visible to the ground target party is "medium"; if z∈[0.25,0.5), the leakage hazard level of the privacy information visible to the ground target party is "relatively minor"; if z∈[0.5,1], the leakage hazard level of the privacy information visible to the ground target party is "minor".

[0057] In one embodiment of the present invention, step S3.1 specifically includes: using detection equipment to monitor whether there is a suspicious drone in the surrounding airspace of the ground target; if the monitoring result shows that there is a suspicious drone, then set p t=1, and record the location of the suspicious drone; if the monitoring result shows that there is no suspicious drone, then set p t =0; if p t =1, then continue to execute steps S3.2 to S3.4; if p t = 0, then let q t =0, and skip steps S3.2 and S3.3, and directly execute step S3.4;

[0058] The step S3.2 specifically includes: Step S3.21: Determine the shooting angle of the nth suspicious drone on the ground target identification marker information surface in, is the horizontal shooting angle, is the vertical shooting angle, and the ground target party identity identification marker information surface is confirmed by the risk assessment party; Step S3.22: by measuring the brightness δ of the ground target party identity identification marker information surface t , the occlusion ratio γ of the ground target identification marker information surface t , environmental visibility level κ t ; where κ t ∈[0,4],γ t ∈[0,1]; Step S3.23: According to the type of ground target identification marker, select and use the pre-trained ground target identification probability evaluation neural network model to evaluate the ground target identification probability p obtained by the nth suspicious drone t n ; Step S3.24: At time point t, the formula corresponding to the probability of the suspicious drone identifying the ground target is:

[0059]

[0060] In formula (3), p t represents the probability of the suspicious drone identifying the ground target at time t; represents the probability of ground target identification obtained by the nth suspicious UAV at time τ; τ represents the evaluation time point identifier, and its value range is all evaluation time points from time 0 to time t; τ∈{0,T 2 ,T 2 *2,T 2 *3...,t};

[0061] Step S3.25: If the ground target does not carry an identification marker, but the risk assessment party infers the probability p of the suspicious drone obtaining the current ground target's identity through existing flight activities based on the existing evidence * , then we can tDirect assignment, that is, p t =p * ;

[0062] The step S3.3 specifically includes: Step S3.31: Determine the shooting angle of the nth suspicious drone on the ground target information plane in, Represents the horizontal shooting angle, represents the vertical shooting angle; the ground target information plane refers to the plane presenting the privacy information of the ground target, which is determined by the risk assessment party; step S3.32: determining the brightness of the ground target by measuring t , ground target information surface occlusion ratio Ambient visibility level κ t ; where κ t ∈[0,4], Step S3.33: According to the type of ground target, select and use the pre-trained ground target visible privacy information clarity assessment neural network model to assess the clarity of the privacy information visible to the ground target photographed by the nth suspicious drone Step S3.34: At time point t, the formula corresponding to the clarity of the private information visible to the ground target photographed by the suspicious drone is obtained as follows:

[0063]

[0064] The step S3.4 specifically includes: Step S3.41: The risk assessment party estimates the estimated length T of the process of generating the visible privacy information of the ground target party according to the real-time situation 1 ; Step S3.42: Based on the estimated length T 1 And formula (5) to calculate the leakage probability of private information visible to the ground target;

[0065]

[0066] In formula (5), is the estimated number of monitoring time points included in the process of generating visible privacy information; T 2 represents the time interval of real-time evaluation; P t represents the probability of leakage of private information visible to the ground target at time t; mm is a variable representing an integer, and its value range is from 0 to t / T 2 Integer: mm∈{0,1,...,t / T 2};

[0067] And the step S3.5 specifically includes: step S3.51: according to the calculation result of the step S3.42, at time point t, the leakage probability level of the visible privacy information of the ground target is evaluated for the fth time in real time; if P t ∈[0,0.1), then the fth real-time evaluation result of the probability level of leakage of the visible privacy information of the ground target is “slight”; if P t ∈[0.1,0.4), then the ground target can see that the fth real-time evaluation result of the probability level of leakage of privacy information is “relatively slight”; if P t ∈[0.4,0.6), then the fth real-time evaluation result of the leakage probability level of the visible privacy information of the ground target is “medium”; if P t ∈[0.6,0.9), the ground target can see that the fth real-time evaluation result of the probability level of privacy information leakage is “more serious”; if P t ∈[0.9,1], the f-th real-time evaluation result of the probability level of leakage of visible privacy information to the ground target is “serious”.

[0068] In one embodiment of the present invention, according to the evaluation result of the leakage hazard level of the privacy information visible to the ground target party obtained in step S1.41 and the f-th real-time evaluation result of the leakage probability level of the privacy information visible to the ground target party obtained in step S3.51, and according to the risk matrix shown in Table 1, the leakage risk level of the privacy information visible to the ground target party is evaluated for the f-th time in real time at time point t;

[0069] Table 1 Risk Matrix

[0070] Specific embodiment one:

[0072] The following will use a specific embodiment to demonstrate the real-time assessment method of the risk of privacy information leakage visible to ground targets caused by a suspicious drone of the present invention.

[0073] The steps for implementing the method for real-time assessment of the risk of privacy information leakage visible to ground targets caused by suspicious drones in this specific embodiment are as follows:

[0074] (1) Step S1: Evaluate the level of leakage risk of the private information visible to the ground target party.

[0075] The step S1 comprises: step S1.1: constructing a network model of the organization where the person related to the ground target party is located; wherein the person related to the ground target party refers to a person associated with the ground target party; the person related to the ground target party is confirmed by the risk assessment party; step S1.2: calculating the importance of each member in the organization where the person related to the ground target party is located based on the network model; step S1.3: calculating the ranking ratio value of the importance of the person related to the ground target party based on the calculation result of the importance of each member in the organization where the person related to the ground target party is located; step S1.4: determining the leakage hazard level of the visible privacy information of the ground target party based on the calculation result of the ranking ratio value of the importance of the person related to the ground target party.

[0076] The step S1.1 specifically includes:

[0077] Step S1.11: The risk assessment party determines the organization of the person related to the ground target party based on the daily social behavior of the person related to the ground target party; the network model of the organization of the person related to the ground target party is constructed as a two-way complex network; wherein each member of the organization of the person related to the ground target party is modeled as a two-way complex network node; Step S1.12: statistics on the information interaction between internal members of the organization of the person related to the ground target party within a certain time span; if member i sends information to member j, an edge from node i to node j is established, and the number of sent information is used as the weight W of the edge ij ; Step S1.13: Establish a super node, establish edges from the super node to other nodes in the bidirectional complex network, and set the weights of the edges to 1; Step S1.14: Set the initial value of the importance S value of each node in the bidirectional complex network to 1;

[0078] The step S1.2 specifically includes:

[0079] Step S1.21: Iteratively calculate the importance S value of each node. When the importance S value of each node converges, the iterative calculation ends. The formula corresponding to the iterative calculation is:

[0080]

[0081] In formula (1), S j (tt) represents the importance of node j calculated at the (tt)th iteration; S i (tt+1) represents the importance of node i calculated at the (tt+1)th iteration; the S value corresponding to each member node is the importance of each member in the organization where the person related to the ground target party is located; m represents the total number of edges with node i as the end point in the network model; W ji W represents the weight of the edge with starting point j and end point i; jprepresents the weight of the edge with starting point j and end point p; q represents the total number of edges with node j as the starting point in the network model; p represents the end point of the edge with node j as the starting point in the network model;

[0082] The step S1.3 specifically includes:

[0083] Step S1.31: According to the calculation result of step S1.21, the importance of each member in the organization of the ground target party is arranged in descending order; Step S1.32: The ranking number of the ground target party is recorded as l, and the total number of organizational personnel is recorded as L, and the ranking ratio value of the importance of the ground target party is calculated; the formula for calculating the ranking ratio value is:

[0084]

[0085] And the step S1.4 specifically includes:

[0086] Step S1.41: According to the calculation result z of step S1.32, determine the leakage hazard level of the privacy information visible to the ground target party; if z∈[0,0.06), the leakage hazard level of the privacy information visible to the ground target party is "serious"; if z∈[0.06,0.125), the leakage hazard level of the privacy information visible to the ground target party is "relatively serious"; if z∈[0.125,0.25), the leakage hazard level of the privacy information visible to the ground target party is "medium"; if z∈[0.25,0.5), the leakage hazard level of the privacy information visible to the ground target party is "relatively minor"; if z∈[0.5,1], the leakage hazard level of the privacy information visible to the ground target party is "minor".

[0087] (2) Step S2: Record the start time of the process of generating the private information visible to the ground target as time point 0, and determine the time interval T for real-time evaluation 2 ; Then every T 2 , that is, at time point t, step S3 and step S4 are executed in sequence; t = (f-1) × T 2 ; f represents the number of real-time evaluations, and f is a positive integer greater than or equal to 1.

[0088] (3) Step S3: At time point t, the leakage probability level of the private information visible to the ground target party is evaluated in real time for the fth time.

[0089] The step S3 includes: step S3.1: at time point t, monitor the surrounding airspace of the ground target party, and according to the monitoring results, execute steps S3.2 to S3.4, or skip steps S3.2 and S3.3 and directly execute step S3.4; step S3.2: at time point t, calculate the probability of the suspicious drone identifying the identity of the ground target party; step S3.3: at time point t, calculate the clarity of the private information visible to the ground target party photographed by the suspicious drone; step S3.4: at time point t, calculate the leakage probability of the private information visible to the ground target party; step S3.5: at time point t, perform the fth real-time evaluation of the leakage probability level of the private information visible to the ground target party.

[0090] The step S3.1 specifically includes:

[0091] Use detection equipment to monitor whether there are suspicious drones in the surrounding airspace of the ground target; if the monitoring result shows that there are suspicious drones, let p t =1, and record the location of the suspicious drone; if the monitoring result shows that there is no suspicious drone, then set p t =0; if p t =1, then continue to execute steps S3.2 to S3.4; if p t = 0, then let q t =0, and skip steps S3.2 and S3.3, and directly execute step S3.4;

[0092] The step S3.2 specifically includes:

[0093] Step S3.21: Determine the shooting angle of the nth suspicious drone on the ground target identification marker information surface in, is the horizontal shooting angle, is the vertical shooting angle, and the ground target party identity identification marker information surface is confirmed by the risk assessment party; Step S3.22: by measuring the brightness δ of the ground target party identity identification marker information surface t , the occlusion ratio γ of the ground target identification marker information surface t , environmental visibility level κ t ; where κ t ∈[0,4],γ t ∈[0,1]; Step S3.23: According to the type of ground target identification marker, select and use the pre-trained ground target identification probability evaluation neural network model to evaluate the ground target identification probability obtained by the nth suspicious drone Step S3.24: At time point t, the formula corresponding to the probability of the suspicious drone identifying the ground target is:

[0094]

[0095] In formula (3), p t represents the probability of the suspicious drone identifying the ground target at time t; represents the probability of ground target identification obtained by the nth suspicious UAV at time τ; τ represents the evaluation time point identifier, and its value range is all evaluation time points from time 0 to time t; τ∈{0,T 2 ,T 2 *2,T 2 *3...,t};

[0096] Step S3.25: If the ground target does not carry an identification marker, but the risk assessment party infers the probability p of the suspicious drone obtaining the current ground target's identity through existing flight activities based on the existing evidence * , then we can t Direct assignment, that is, p t =p * ; The step S3.3 specifically includes: Step S3.31: Determine the shooting angle of the nth suspicious drone on the ground target information surface in, Represents the horizontal shooting angle, represents the vertical shooting angle; the ground target information plane refers to the plane presenting the privacy information of the ground target, which is determined by the risk assessment party; step S3.32: determining the brightness of the ground target by measuring t , ground target information surface occlusion ratio Ambient visibility level κ t ; where κ t ∈[0,4], Step S3.33: According to the type of ground target, select and use the pre-trained ground target visible privacy information clarity assessment neural network model to assess the clarity of the privacy information visible to the ground target photographed by the nth suspicious drone Step S3.34: At time point t, the formula corresponding to the clarity of the private information visible to the ground target photographed by the suspicious drone is obtained as follows:

[0097]

[0098] The step S3.4 specifically includes: Step S3.41: The risk assessment party estimates the estimated length T of the process of generating the visible privacy information of the ground target party according to the real-time situation 1 ; Step S3.42: Based on the estimated length T 1And formula (5) to calculate the leakage probability of private information visible to the ground target;

[0099]

[0100] In formula (5), is the estimated number of monitoring time points included in the process of generating visible privacy information; T 2 represents the time interval of real-time evaluation; P t represents the probability of leakage of private information visible to the ground target at time t; mm is a variable representing an integer, and its value range is from 0 to t / T 2 Integer: mm∈{0,1,...,t / T 2};

[0101] And the step S3.5 specifically includes:

[0102] Step S3.51: Based on the calculation result of step S3.42, at time point t, the probability level of leakage of the visible private information of the ground target is evaluated for the fth time in real time; if P t ∈[0,0.1), then the fth real-time evaluation result of the probability level of leakage of the visible privacy information of the ground target is “slight”; if P t ∈[0.1,0.4), then the ground target can see that the fth real-time evaluation result of the probability level of leakage of privacy information is “relatively slight”; if P t ∈[0.4,0.6), then the fth real-time evaluation result of the leakage probability level of the visible privacy information of the ground target is “medium”; if P t ∈[0.6,0.9), the ground target can see that the fth real-time evaluation result of the probability level of privacy information leakage is “more serious”; if P t ∈[0.9,1], the f-th real-time evaluation result of the probability level of leakage of visible privacy information to the ground target is “serious”.

[0103] Specifically, the definition process of the horizontal shooting angle and the vertical shooting angle in step S3.21 is as follows: assuming that the shooting target information plane is O, the shooting target center point on the information plane O is o, and the shooting point is u; when the plane O is perpendicular to the horizontal plane, construct a horizontal line l passing through point o 1 , construct a ray l that passes through point o and is vertically upward 2 , l 1 and l 2 The trajectories on O are l 1 'and l' 2 , after l 1 ' and the plane perpendicular to O is O 2 , after l' 2The plane perpendicular to O is O 3 ; In plane O 2 With O 3 On the intersection line, construct a ray l with point o as the starting point and direction as the front direction of the shooting target 3 ; The straight line passing through o and u is l 4 ; then l 4 In O 2 The projection on l 3 The clockwise angle of is the horizontal shooting angle, l 4 In O 3 The projection on l 3 The angle between the vertical direction and the horizontal direction is the vertical shooting angle.

[0104] Specifically, in step S3.21 and step S3.31, the specific method for calculating the shooting angle of the suspicious drone to the shooting target information plane is as follows: set the radar coordinate system and the shooting target coordinate system, wherein the radar coordinate system takes the center point of the monitoring device as the origin, and the shooting target coordinate system takes the center of the shooting target as the origin, and both coordinate systems take the due north direction as the positive direction of the x-axis, the due east direction as the positive direction of the y-axis, and the vertical upward direction as the z-axis. The position information of the measuring point in the coordinate system is expressed in the format of (R, α, β), wherein R represents the slant distance, that is, the distance between the measuring point and the origin, α represents the azimuth of the measuring point, which refers to the clockwise angle between the projection line of the slant distance on the horizontal plane passing through the x-axis and the y-axis and the x-axis, and its value range is 0 to 360 degrees, and β represents the pitch angle of the measuring point, which refers to the angle between the projection line of the slant distance on the vertical plane passing through the x-axis and the z-axis and the x-axis, and its value range is -90 degrees to +90 degrees.

[0105] The airspace monitoring device obtains the position of the nth UAV in the radar coordinate system as (R n ,α n ,β n ), the position of the shooting target is (R 0 ,α 0 ,β 0 ), the method for calculating the position information of the nth UAV in the target coordinate system is as follows:

[0106] The coordinates on the z-axis are: The coordinate on the x-axis is

[0107] The coordinate on the y-axis is The azimuth is: The pitch angle is:

[0108] The direction of the target information surface based on the target coordinate system is expressed as (α ** ,β **) indicates that, assuming that ray l is a ray starting from the center of the photographed target, extending in the direction of the front of the information surface and perpendicular to the photographed target information surface, α ** is the azimuth of the point on ray l, β ** is the pitch angle of the point on the ray.

[0109] The specific calculation method of the shooting angle of the nth suspicious drone to the target information surface is: the horizontal shooting angle σ n for:

[0110]

[0111] Vertical shooting angle

[0112] Specifically, in step S3.23, the specific method for pre-training the neural network model for probability assessment of identity recognition of various types of ground targets is: constructing a data set, building a neural network, training and verifying the neural network. Among them, the specific method for constructing the data set is: collecting photos of a certain type of ground target identity marker under different shooting angles, different ground target identity marker information surface brightness, different ground target identity marker information surface occlusion ratio, and different environmental visibility levels. The horizontal shooting angle, the vertical shooting angle, the ground target identity marker information surface brightness, the ground target identity marker information surface occlusion ratio, and the environmental visibility level are used as input. The similarity between the photographed photo and the front photo of the ground target identity marker information surface is used as the output. The specific method for building the neural network is: the neural network includes an input layer, 2 hidden layers and an output layer, wherein the input layer contains 5 nodes, the hidden layer contains 64 nodes, and the output layer contains 1 node. The neural network is a fully connected network, and the activation function of each node in the hidden layer is the ReLU function, and the output layer activation function uses the Sigmoid function. The specific method of training and verifying the neural network is: divide the data set into training set, test set, and validation set in a ratio of 7:1:2; the Adam algorithm is used for model training. In the specific implementation process, MAE (mean absolute error) is used as the loss function, and the Adam algorithm is used to train the neural network. The learning rate is set to 0.001, the batch size is set to 64, and a total of 100 rounds of training are performed. The hardware platform processor used is AMD Ryzen5600, and the GPU is NVIDIA RTX3060.

[0113] Specifically, in step S3.33, the specific method for pre-training the neural network model for assessing the clarity of privacy information visible to each type of ground target is: constructing a data set, building a neural network, training and verifying the neural network. Among them, the specific method for constructing the data set is: collecting photos of a certain type of ground target under different shooting angles, different ground target brightness, different ground target information surface occlusion ratios, and different environmental visibility levels. The horizontal shooting angle, vertical shooting angle, ground target brightness, ground target information surface occlusion ratio, and environmental visibility level are used as input. Experts annotate the photos taken, and the annotation content is a clarity score, with a score range of 0 to 1, 0 means completely unclear, 1 means completely clear, and the annotation result is used as output. The specific method for building a neural network is: the neural network includes an input layer, 2 hidden layers, and an output layer, wherein the input layer contains 5 nodes, the hidden layer contains 64 nodes, and the output layer contains 1 node. The neural network is a fully connected network, and the activation function of each node in the hidden layer is the ReLU function, and the output layer activation function uses the Sigmoid function. The specific method of training and verifying the neural network is: divide the data set into training set, test set, and validation set in a ratio of 7:1:2; the Adam algorithm is used for model training. In the specific implementation process, MAE (mean absolute error) is used as the loss function, and the Adam algorithm is used to train the neural network. The learning rate is set to 0.001, the batch size is set to 64, and a total of 100 rounds of training are performed. The hardware platform processor used is AMD Ryzen 5600, and the GPU is NVIDIA RTX3060.

[0114] (4) Step S4: Based on the evaluation result of the leakage hazard level of the privacy information visible to the ground target party and the f-th real-time evaluation result of the leakage probability level of the privacy information visible to the ground target party, the f-th real-time evaluation of the leakage risk level of the privacy information visible to the ground target party is performed at time point t until the generation process of the privacy information visible to the ground target party is completed and the evaluation is stopped.

[0115] Further, the step S4 includes: performing an f-th real-time evaluation of the leakage risk level of the ground target party visible privacy information at time point t according to the evaluation result of the leakage hazard level of the ground target party visible privacy information obtained in step S1.41 and the f-th real-time evaluation result of the leakage probability level of the ground target party visible privacy information obtained in step S3.51, and according to the risk matrix shown in Table 1;

[0116] Table 1 Risk Matrix

[0117] Specific embodiment 2:

[0119] The following is a specific embodiment of the present invention to demonstrate the real-time assessment system for the risk of privacy information leakage visible to the ground target party caused by a suspicious drone. Based on the same inventive idea of ​​the real-time assessment method for the risk of privacy information leakage visible to the ground target party caused by a suspicious drone in the specific embodiment 1, the specific embodiment 2 discloses a real-time assessment system for the risk of privacy information leakage visible to the ground target party caused by a suspicious drone, including: a ground target party visible privacy information leakage hazard level assessment module, used to calculate the importance of each member in the organization where the ground target party-related person is located, and further assess the ground target party visible privacy information leakage hazard level; a ground target party visible privacy information leakage probability level assessment module, used to estimate the probability of the suspicious drone identifying the ground target party and the clarity of the ground target party visible privacy information photographed by the drone, and further determine the ground target party visible privacy information leakage probability level; a ground target party visible privacy information leakage risk level assessment module, used to determine the ground target party visible privacy information leakage risk level according to the ground target party visible privacy information leakage hazard level and the ground target party visible privacy information leakage probability level.

[0120] In summary, the key technical points of the present invention are as follows: 1. The present invention evaluates the risk of leakage of private information visible to ground targets caused by suspicious drones, that is, the risk level is evaluated in real time according to the probability of leakage of visible private information and the degree of hazard of leakage of visible private information. 2. The present invention evaluates the degree of hazard of leakage of private information visible to ground targets. 3. The present invention calculates the probability of leakage of private information visible to ground targets. 4. The present invention calculates the probability of identification of the identity of the ground target by a suspicious drone. 5. The present invention calculates the clarity of private information visible to ground targets photographed by a suspicious drone. 6. The present invention calculates the shooting angle of the drone for the target information surface. 7. The present invention constructs and trains a neural network model for evaluating the probability of identification of the ground target. 8. The present invention constructs and trains a neural network model for evaluating the clarity of private information visible to the ground target.

[0121] Therefore, it is proved that the risk assessment method and system for privacy information leakage visible to ground targets caused by suspicious drones provided by the present invention is completely a research gap, and the assessment method has good rationality and feasibility.

[0122] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A real-time assessment method for the risk of private information leakage visible to ground targets caused by suspicious drones, comprising: Step S1: Evaluate the level of leakage risk of the private information visible to the ground target party; Step S2: Record the start time of the process of generating the private information visible to the ground target party as time point 0, and determine the time interval T2 of the real-time evaluation; then, every T2, that is, at time point t, run step S3 and step S4 in sequence; t = (f-1) × T2; f represents the number of real-time evaluations, and f is a positive integer greater than or equal to 1; Step S3: at time point t, the leakage probability level of the private information visible to the ground target party is evaluated in real time for the fth time; Step S4: Based on the evaluation result of the leakage hazard level of the privacy information visible to the ground target party and the f-th real-time evaluation result of the leakage probability level of the privacy information visible to the ground target party, perform the f-th real-time evaluation of the leakage risk level of the privacy information visible to the ground target party at time point t until the generation process of the privacy information visible to the ground target party is completed and the evaluation is stopped.

2. The method for real-time assessment of the risk of privacy information leakage visible to ground targets caused by suspicious drones according to claim 1 is characterized in that: The step S1 comprises: Step S1.1: construct a network model of the organization where the ground target party related person is located; wherein the ground target party related person refers to a person associated with the ground target party; the ground target party related person is confirmed by the risk assessment party; Step S1.2: based on the network model, calculate the importance of each member in the organization where the ground target party related person is located; Step S1.3: Calculate the ranking ratio value of the importance of the ground target party related person based on the calculation result of the importance of each member in the organization to which the ground target party related person belongs; Step S1.4: Based on the calculation result of the ranking ratio value of the importance of the ground target party related persons, determine the leakage hazard level of the ground target party's visible privacy information.

3. The method for real-time assessment of the risk of privacy information leakage visible to ground targets caused by suspicious drones according to claim 2 is characterized in that: The step S3 comprises: Step S3.1: At time point t, monitor the surrounding airspace of the ground target, and according to the monitoring results, execute steps S3.2 to S3.4, or skip steps S3.2 and S3.3 and directly execute step S3.4; Step S3.2: At time point t, calculate the probability of the suspicious drone identifying the ground target party; Step S3.3: at time point t, calculate the clarity of the private information visible to the ground target photographed by the suspicious drone; Step S3.4: at time point t, calculate the leakage probability of the private information visible to the ground target; Step S3.5: At time point t, the leakage probability level of the private information visible to the ground target party is evaluated in real time for the fth time.

4. The method for real-time assessment of the risk of privacy information leakage visible to ground targets caused by suspicious drones according to claim 3 is characterized in that: The step S1.1 specifically includes: Step S1.11: The risk assessment party determines the organization of the person related to the ground target party according to the daily social behavior of the person related to the ground target party; the network model of the organization of the person related to the ground target party is constructed as a two-way complex network; wherein each member of the organization of the person related to the ground target party is modeled as a two-way complex network node; Step S1.12: Statistics are collected on the information interaction between internal members of the organization of the ground target party related person within a certain time span; if member i sends information to member j, an edge is established from node i to node j, and the number of sent information is used as the weight W of the edge. ij ; Step S1.13: Establish a super node, establish edges from the super node to other nodes in the bidirectional complex network, and set the weights of the edges to 1; Step S1.14: Set the initial value of the importance S of each node in the bidirectional complex network to be 1; The step S1.2 specifically includes: Step S1.21: Iteratively calculate the importance S value of each node. When the importance S value of each node converges, the iterative calculation ends. The formula corresponding to the iterative calculation is: In formula (1), S j (tt) represents the importance of node j calculated at the (tt)th iteration; S i (tt+1) represents the importance of node i calculated at the (tt+1)th iteration; the S value corresponding to each member node is the importance of each member in the organization where the person related to the ground target party is located; m represents the total number of edges with node i as the end point in the network model; W ji W represents the weight of the edge with starting point j and end point i; jp represents the weight of the edge with starting point j and end point p; q represents the total number of edges with node j as the starting point in the network model; p represents the end point of the edge with node j as the starting point in the network model; The step S1.3 specifically includes: Step S1.31: According to the calculation result of step S1.21, the importance of each member in the organization of the ground target party related person is arranged in descending order; Step S1.32: record the ranking number of the ground target party related person as l, record the total number of organization personnel as L, and calculate the ranking ratio value of the importance of the ground target party related person; the formula for calculating the ranking ratio value is: And the step S1.4 specifically includes: Step S1.41: According to the calculation result z of step S1.32, determine the leakage hazard level of the privacy information visible to the ground target party; if z∈[0,0.06), the leakage hazard level of the privacy information visible to the ground target party is "serious"; if z∈[0.06,0.125), the leakage hazard level of the privacy information visible to the ground target party is "relatively serious"; if z∈[0.125,0.25), the leakage hazard level of the privacy information visible to the ground target party is "medium"; if z∈[0.25,0.5), the leakage hazard level of the privacy information visible to the ground target party is "relatively minor"; if z∈[0.5,1], the leakage hazard level of the privacy information visible to the ground target party is "minor".

5. The method for real-time assessment of the risk of privacy information leakage visible to ground targets caused by suspicious drones according to claim 4 is characterized in that: The step S3.1 specifically includes: Use detection equipment to monitor whether there are suspicious drones in the surrounding airspace of the ground target; if the monitoring result shows that there are suspicious drones, let p t =1, and record the location of the suspicious drone; if the monitoring result shows that there is no suspicious drone, then set p t =0; if p t =1, then continue to execute steps S3.2 to S3.4; if p t = 0, then let q t =0, and skip steps S3.2 and S3.3, and directly execute step S3.4; The step S3.2 specifically includes: Step S3.21: Determine the shooting angle of the nth suspicious drone on the ground target identification marker information surface in, is the horizontal shooting angle, is a vertical shooting angle, and the ground target party identity identification marker information surface is confirmed by the risk assessment party; Step S3.22: Obtain the brightness δ of the ground target identification marker information surface by measuring t , the occlusion ratio γ of the ground target identification marker information surface t , environmental visibility level κ t ; where κ t ∈[0,4],γ t ∈[0,1]; Step S3.23: Based on the type of ground target identification marker, select and use the pre-trained ground target identification probability evaluation neural network model to evaluate the ground target identification probability obtained by the nth suspicious drone. Step S3.24: At time point t, the formula corresponding to the probability of the suspicious drone identifying the ground target is: In formula (3), p t represents the probability of the suspicious drone identifying the ground target at time t; represents the probability of ground target identification obtained by the nth suspicious drone at time τ; τ represents the evaluation time point identifier, and its value range is all evaluation time points from time 0 to time t; τ∈{0,T2,T2*2,T2*3...,t}; Step S3.25: If the ground target does not carry an identification marker, but the risk assessment party infers the probability p of the suspicious drone obtaining the current ground target's identity through existing flight activities based on the existing evidence * , then we can t Direct assignment, that is, p t =p * ; The step S3.3 specifically includes: Step S3.31: Determine the shooting angle of the nth suspicious drone on the ground target information plane in, Represents the horizontal shooting angle, represents the vertical shooting angle; the ground target information plane refers to the plane presenting the privacy information of the ground target, which is determined by the risk assessment party; Step S3.32: Determine the ground target brightness ρ by measuring t , ground target information surface occlusion ratio Ambient visibility level κ t ; where κ t ∈[0,4], Step S3.33: According to the type of ground target, select and use the pre-trained ground target visible privacy information clarity assessment neural network model to assess the clarity of the privacy information visible to the ground target photographed by the nth suspicious drone Step S3.34: At time point t, the formula corresponding to the clarity of the private information visible to the ground target photographed by the suspicious drone is obtained as follows: The step S3.4 specifically includes: Step S3.41: The risk assessment party estimates the estimated length T1 of the process of generating the visible private information of the ground target party according to the real-time situation; Step S3.42: Based on the estimated length T1 and formula (5), the leakage probability of the private information visible to the ground target is calculated; In formula (5), is the estimated number of monitoring time points included in the process of generating visible privacy information; T2 represents the time interval of real-time evaluation; P t represents the leakage probability of the private information visible to the ground target at time t; mm is a variable representing an integer, and its value range is an integer from 0 to t / T2: mm∈{0,1,...,t / T2}; And the step S3.5 specifically includes: Step S3.51: Based on the calculation result of step S3.42, at time point t, the leakage probability level of the private information visible to the ground target party is evaluated in real time for the fth time; If P t ∈[0,0.1), the fth real-time evaluation result of the probability level of leakage of the visible privacy information of the ground target is "slight"; if P t ∈[0.1,0.4), then the ground target can see that the fth real-time evaluation result of the probability level of leakage of privacy information is "relatively slight"; if P t ∈[0.4,0.6), then the fth real-time evaluation result of the probability level of leakage of the visible privacy information of the ground target is "medium"; if P t ∈[0.6,0.9), the ground target party can see that the fth real-time evaluation result of the probability level of leakage of privacy information is "more serious"; if P t ∈[0.9,1], the f-th real-time evaluation result of the probability level of leakage of the visible privacy information of the ground target is "serious".

6. The method for real-time assessment of the risk of privacy information leakage visible to ground targets caused by suspicious drones according to claim 5 is characterized in that: The step S4 comprises: performing an f-th real-time evaluation of the leakage risk level of the ground target party visible privacy information at time point t according to the evaluation result of the leakage hazard level of the ground target party visible privacy information obtained in step S1.41 and the f-th real-time evaluation result of the leakage probability level of the ground target party visible privacy information obtained in step S3.51, and according to the risk matrix shown in Table 1; Table 1 Risk Matrix

Images