Civil aviation passenger security check verification method based on multi-source heterogeneous data real-time fusion

Abstract

The invention discloses a civil aviation passenger security check verification method based on multi-source heterogeneous data real-time fusion, and the method comprises the steps: carrying out the format conversion, data coding unification, field name unification and timestamp of data from different source systems at the rear end of a security check information system, and carrying out the data connection, projection and dimension reduction combination, thereby obtaining a security check information system; all data from different systems are merged into a to-be-verified passenger data table, index conditions are established according to travel information such as passenger certificate numbers, flight numbers, flight dates, departure stations and terminal buildings, and at the front end of a security check information system, only basic travel information of passengers needs to be used as indexes during passenger verification, so that the verification efficiency is improved. According to the method, all information needing to be verified of the passenger can be quickly obtained and verified by only needing to carry out verification, the security check efficiency is effectively improved, data fusion is carried out factually, the accuracy and the real-time performance of security check verification are ensured, and the problems that data of an existing security check information system are incomplete and intersection is complex are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of airport security inspection, and in particular relates to a civil aviation passenger security inspection verification method based on real-time fusion of multi-source heterogeneous data. Background Art

[0002] Before entering the airport security check area, passengers need to be verified at the security check counter. The verification content includes passenger identity information, flight information, flight status, control information, checked baggage unpacking information, etc. The traditional verification solution is to pre-process the data pushed by each source system and store it in the corresponding database table. The information that passengers need to verify for security check is scattered in various database tables; when passengers go through security check, the verification method used is to query data from each database in turn through the passenger's ID number, and verify the passenger's basic information, check-in status, flight status, luggage unpacking status, whether there are control personnel, etc. in turn, which makes the security check slow and inefficient.

[0003] The system responsible for verification is hereinafter referred to as the security information system. The existing security information system faces the following problems: 1. Data heterogeneity: Data formats, protocols, and encoding methods from different systems vary, making it difficult to quickly integrate; 2. Incomplete data: Information from a single data source may not be sufficient to support comprehensive identity verification, posing potential security risks; 3. Complex data intersections: There may be duplication or conflict between multi-source data, resulting in redundant or contradictory information, which increases the difficulty and time cost of verification. Summary of the invention

[0004] The purpose of the present invention is to provide a civil aviation passenger security verification method based on real-time fusion of multi-source heterogeneous data. After data preprocessing, real-time data fusion is added to merge all information that needs to be verified into a data table to be verified. When verifying passengers, only the basic itinerary information of the passengers is needed as an index to quickly obtain and verify all the information that needs to be verified by the passengers, thereby ensuring the accuracy and real-time nature of security verification and improving security efficiency.

[0005] To achieve the above object, the present invention provides a civil aviation passenger security verification method based on real-time fusion of multi-source heterogeneous data, comprising the following steps: S1. At the back end of the security information system, data from different source systems are integrated to form a passenger data table to be verified. Specifically, the following steps are included: S1.1. Input data: Each source system pushes original data to the security information system, including passenger booking data, passenger check-in data, flight schedule data, flight dynamic data, baggage security data, baggage transportation data, special passenger data, and restricted list passenger data; S1.2, Data preprocessing: Convert the data from different source systems into different formats, unify the data coding, field names and timestamps; S1.3, basic information integration: using data connection and projection, passenger booking data and check-in data are connected through passenger ID number, and the required basic passenger information is filtered into the passenger data table to be verified. The basic passenger information includes passenger ID number, flight number, flight date, departure station, terminal, and check-in status; S1.4, Flight data fusion: Merge flight dynamic data and flight plan data, select the latest flight status information to the passenger data table to be verified, and the flight status information can show whether the flight is normal or abnormal; S1.5, Baggage data fusion: Merge the baggage status information in the baggage security inspection data and the baggage transportation data, obtain the status information of whether the checked baggage needs to be unpacked and put it into the passenger data table to be verified; S1.6, Control data fusion: Search the restricted list passenger data and special passenger data by the passenger ID number to determine whether control is needed, and obtain the control information to the passenger data table to be verified. The control information includes whether control is needed and the reason for control; S2. Passenger security verification is performed at the front end of the security information system, including the following steps: S2.1. Verify the passenger's identity information: collect the passenger's ID photo information and compare it with the passenger's face information collected by the real-time camera to verify the consistency between the passenger and the ID; S2.2. Verify the basic information of the passenger: extract the passenger's itinerary information from the boarding pass provided by the passenger, the passenger's itinerary information includes the passenger's ID number, flight number, flight date, departure station, and terminal. Using the passenger's itinerary information as the index condition, obtain the passenger's verification status information from the passenger data table to be verified for security verification. The verification status information includes whether the passenger has checked in, whether the flight status is normal, whether the checked baggage has been unpacked, and whether control has been performed, so as to prompt the security operator whether to unpack and perform control operations.

[0006] Further, in step S1.3, the formula for generating the passenger basic information Passenger_Info is as follows:

[0007] Among them, π is a projection operation, which means retaining specific items; σ is a selection operation, which is used to filter data that meets the conditions; Passenger_ID, Flight_Number, Flight_Date, Departure_Station, Depature_Terminal, and Is_Checked represent passenger ID number, flight number, flight date, departure station, terminal, and check-in status, respectively; It is a natural join operator, used for natural join operation; Seat_Data is passenger reservation data, and CheckIn_Data is check-in data.

[0008] Further, in step S1.4, the formula for obtaining the flight status information Flight_Status is: Flight_Status=π is_normal (σ latest (Flight_Dynamic ∪ Flight_Scheduled) ) Among them, π is_normal To retain flight status information and determine whether the flight status is normal; latest Indicates selecting the latest data based on timestamp; Flight_Dynamic is flight dynamic data; Flight_Scheduled is flight schedule data; ∪ indicates a merge operation.

[0009] Further, in step S1.4, if the filtered flight status information is planned, check-in started, boarding started, or arrived, the flight is in a normal state; if the filtered flight status information is canceled, boarding is closed, or taken off, the flight is in an abnormal state.

[0010] Further, in step S1.5, the baggage status information after combining the baggage security inspection data and the baggage transportation data includes two status information: abnormality of the baggage and no abnormality of the baggage. If the baggage status information contains information that the baggage is in an abnormal state, it is determined that the baggage needs to be unpacked. If the baggage status information does not contain information that the baggage is in an abnormal state, it is determined that the baggage does not need to be unpacked. The formula for obtaining the status information Baggage_Status of whether the checked baggage needs to be unpacked is:

[0011] Among them, π Passenger_ID,Baggage_Check_Required It indicates the operation of judging whether the baggage is unpacked and retaining the information of whether the baggage is unpacked and the passenger's ID number; σ is a selection operation, which is used to filter the data that meets the conditions; Baggage_Abnormal=True indicates that the baggage has an abnormality; Baggage_Abnormal=False indicates that the baggage has no abnormality; Baggage_Data indicates the baggage status information after merging the baggage security inspection data and the baggage transportation data; ∪ indicates a merge operation.

[0012] Further, in step S1.5, if the luggage security inspection data records information about prohibited items abnormalities, over-specification abnormalities, and overweight abnormalities, then the luggage is abnormal and needs to be unpacked; if the luggage transportation data records information about damaged luggage, then the luggage is abnormal and needs to be unpacked.

[0013] Further, in step S1.6, firstly, according to the passenger ID number Passenger_ID, the passenger data Restric_Data on the restricted list is retrieved to see whether the passenger is on the restricted list. If yes, the passenger is determined to be a controlled person and the control reason is obtained from the passenger data on the restricted list. Then, the control information is retained in the passenger data table to be verified. The formula for obtaining the control information on the restricted list is as follows: Restric_Control=π Passenger_ID,Restric_Reason (σ Passenger_ID∈Restric _ Data (Restric_Data)) Among them, Restric_Control represents the restricted list control information; π is a projection operation, indicating that a specific item is retained; Passenger_ID represents the passenger ID number; Restric_Reason represents the reason for being included in the restricted list; σ is a selection operation, used to filter data that meets the conditions; At this time, the restricted list control information Restric_Control is the final control information Final_Control; If the passenger is not in the restricted list, the special passenger data Special_Data is searched. If the passenger is in the special passenger data, the passenger is determined to be a controlled person and the control reason is obtained from the special passenger data to generate the final control information. The final control information is then retained in the passenger data table to be verified. The formula for obtaining the final control information is as follows: If(Restric_Control==null){Final_Control=π Passenger_ID,Special_Reason (σ Passenger_ID∈Special_Data (Special_Reason))} Among them, Restric_Control==null means that the restricted list control information is empty; Special_Reason indicates the reason for being listed as a special passenger; If the passenger is not included in the special passenger data, it is determined that the passenger is not a monitored person and does not need to be monitored.

[0014] Further, in step S2.1, verifying the identity information of the passenger specifically includes the following steps: S2.1.1、Reading ID information: When a passenger passes through the security check counter, the ID information of the passenger is read through the scanning device to extract the ID number and ID photo information of the passenger; S2.1.2. Extract ID photo features: Use image processing algorithms to extract facial feature points from the ID photo information to form a digital facial feature vector; S2.1.3, Face recognition extraction: The camera installed at the security checkpoint collects the passenger's current face image in real time, and uses the image processing algorithm to extract the passenger's facial features from the face image collected by the camera to generate a face feature vector; S2.1.4, Feature comparison: The facial feature vector extracted from the ID card is compared with the facial feature vector collected in real time by the camera, and the similarity between the two is calculated using the Euclidean distance or cosine similarity. The calculation formula for the Euclidean distance is:

[0015] Among them, f ID represents the facial feature vector extracted from the document, f face Represents the face feature vector captured in real time; i is the sequence number of the feature; Represents the vector value of the ID’s facial image on the i-th feature; Represents the vector value of the passenger’s face photographed on site on the i-th feature; The calculation formula for cosine similarity is:

[0016] A similarity threshold is set. If the calculated cosine similarity is higher than the similarity threshold, it is considered that the ID photo information matches the passenger's current facial information, the verification is passed, and the next verification step is allowed; otherwise, the verification fails, and the security information system prompts the operator to perform manual review to further verify the passenger's identity.

[0017] Further, in step S1.1, the source systems include a ticket booking system, a departure system, an airport information integration system, a baggage security inspection system, a baggage sorting system, a Civil Aviation Administration system and a public security system; the passenger reservation data originates from the ticket booking system, the passenger check-in information originates from the departure system, the flight schedule data and flight dynamic data originate from the airport information integration system, the baggage security inspection data originates from the baggage security inspection system, the baggage transportation data originates from the baggage sorting system, the special passenger data originates from the Civil Aviation Administration system, and the restricted list passenger data originates from the public security system.

[0018] After adopting the above scheme, the beneficial effects of the present invention are: During data preprocessing, the present invention converts the format of data from different source systems, unifies data coding, field names and timestamps, and facilitates subsequent data fusion and integration, thereby solving the problem that heterogeneous data in existing security inspection information systems are difficult to integrate.

[0019] More importantly, after data preprocessing, a real-time data fusion step is added. Through data connection, projection and dimensionality reduction, all data from different systems are merged into a passenger data table to be verified, and index conditions are established according to travel information such as passenger ID number, flight number, flight date, departure station, terminal, etc. When verifying passengers, only the basic travel information of the passengers is used as the index to quickly obtain and verify all the information that needs to be verified by the passengers, effectively improving the security inspection efficiency. Moreover, the data fusion is carried out in fact, ensuring the accuracy and real-time nature of security inspection verification, and solving the problem of incomplete data and complex intersections in the existing security inspection information system. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a process framework diagram of step S1 of the present invention; Figure 2 This is a flow chart of step S2.1 of the present invention; Figure 3 This is a flow chart of step S2.2 of the present invention. DETAILED DESCRIPTION

[0021] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] like Figure 1-3 As shown, the present invention provides a civil aviation passenger security verification method based on real-time fusion of multi-source heterogeneous data, comprising the following steps: S1. At the back end of the security information system, data from different source systems are integrated to form a passenger data table to be verified, such as Figure 1 As shown, the specific steps include: S1.1. Input data: Each source system pushes original data to the security information system, including passenger reservation data, passenger check-in data, flight schedule data, flight dynamic data, baggage security inspection data, baggage transportation data, special passenger data, and restricted list passenger data.

[0023] The source systems include the ticket booking system, the departure system, the airport information integration system, the baggage security inspection system, the baggage sorting system, the Civil Aviation Administration system and the public security system. The passenger booking data comes from the ticket booking system. After the passenger orders the ticket online or offline, the ticket booking system can obtain the passenger booking data; the passenger check-in information comes from the departure system, the flight schedule data and flight dynamic data come from the airport information integration system, the baggage security inspection data comes from the baggage security inspection system, and the baggage transportation data comes from the baggage sorting system; the special passenger data comes from the Civil Aviation Administration system. The special passengers are passengers recorded by the Civil Aviation Administration who need special control operations, such as VIP passengers and untrustworthy passengers. VIP passengers need special reception, and untrustworthy passengers are marked by the Civil Aviation Administration due to their bad reputation. Security inspection operators will restrict untrustworthy passengers from boarding; the restricted list passenger data comes from the public security system. The restricted list passengers are generally people who have violated the law and are restricted from traveling, and need to be arrested and controlled.

[0024] S1.2. Data preprocessing: Convert the data from different source systems into different formats, unify data coding, field names and timestamps to ensure that the data is factual and comparable, so as to facilitate subsequent data fusion and integration and solve the problem of heterogeneous data in existing security inspection information systems that are difficult to integrate.

[0025] S1.3, basic information fusion: the passenger booking data includes flight number, flight date, terminal, booking class, aircraft model and other information, the check-in data includes passenger name, passenger ID number, departure station, arrival station and other information. Through data connection and projection, the passenger booking data and check-in data can be connected through the passenger ID number, and the required passenger basic information can be screened into the passenger data table to be verified. The passenger basic information includes passenger ID number, flight number, flight date, departure station, terminal, and check-in status.

[0026] The formula for generating the basic passenger information Passenger_Info is as follows:

[0027] Among them, π is a projection operation, which means retaining specific items; σ is a selection operation, which is used to filter data that meets the conditions; Passenger_ID, Flight_Number, Flight_Date, Departure_Station, Depature_Terminal, and Is_Checked represent passenger ID number, flight number, flight date, departure station, terminal, and check-in status, respectively; It is a natural join operator, used for natural join operation; Seat_Data is passenger reservation data, and CheckIn_Data is check-in data.

[0028] S1.4, Flight data fusion: Merge flight dynamic data and flight plan data. The flight dynamic data refers to the real-time status information of the flight during takeoff, flight and landing, including the estimated take-off time, actual take-off time, delayed or cancelled flights, etc.; the flight plan data refers to the detailed plan information formulated by the airline to operate the flight, including plan status information and check-in time, boarding time, etc.

[0029] After merging the flight dynamic data and the flight plan data, select the latest flight status information to be added to the passenger data table to be verified. The flight status information can show whether the flight is in a normal or abnormal state. If the selected flight status information is planned, check-in started, boarding started, or arrived, the flight is in a normal state; if the selected flight status information is canceled, boarding is closed, or has taken off, the flight is in an abnormal state.

[0030] The formula for obtaining the flight status information Flight_Status is: Flight_Status=π is_normal (σ latest (Flight_Dynamic ∪ Flight_Scheduled) ) Among them, π is_normal To retain flight status information and determine whether the flight status is normal; latest Indicates selecting the latest data based on timestamp; Flight_Dynamic is flight dynamic data; Flight_Scheduled is flight schedule data; ∪ indicates a merge operation.

[0031] S1.5, Baggage data fusion: Merge the baggage status information in the baggage security inspection data and the baggage transportation data, and obtain the status information of whether the checked baggage needs to be unpacked to the passenger data table to be verified.

[0032] Specifically, the baggage status information after combining the baggage security inspection data and the baggage transportation data includes two status information: abnormal baggage and no abnormal baggage. If the baggage status information contains information that the baggage is in an abnormal state, it is determined that the baggage needs to be unpacked. That is, if there is information that the baggage is in an abnormal state in both the baggage security inspection data and the baggage transportation data, or if there is information that the baggage is in an abnormal state in either of the two data, the baggage is abnormal and needs to be unpacked. Only when there is no information that the baggage is in an abnormal state in both data, the baggage status information shows that there is no abnormal baggage and no need to be unpacked.

[0033] The information on abnormal status of luggage includes abnormal information such as prohibited items, oversize, overweight, etc. recorded in the luggage security data; and information on damaged luggage recorded in the luggage transportation data.

[0034] The formula for obtaining the status information Baggage_Status, which indicates whether the checked baggage needs to be unpacked, is:

[0035] Among them, π Passenger_ID,Baggage_Check_Required It indicates the operation of judging whether the baggage is unpacked and retaining the information of whether the baggage is unpacked and the passenger's ID number; σ is a selection operation, which is used to filter the data that meets the conditions; Baggage_Abnormal=True indicates that the baggage has an abnormality; Baggage_Abnormal=False indicates that the baggage has no abnormality; Baggage_Data indicates the baggage status information after merging the baggage security inspection data and the baggage transportation data; ∪ indicates a merge operation.

[0036] S1.6, control data fusion: search the restricted list passenger data and special passenger data by the passenger ID number to determine whether control is needed, and obtain the control information to the passenger data table to be verified. The control information includes whether control is needed and the reason for control.

[0037] Specifically, firstly, according to the passenger ID number Passenger_ID, the passenger data Restric_Data on the restricted list is retrieved to see if the passenger is on the restricted list. If yes, the passenger is determined to be a controlled person and the control reason is obtained from the passenger data on the restricted list. The control reason may be that the passenger violated the law and is restricted from traveling. Then the control information is retained in the passenger data table to be verified. The formula for obtaining the control information on the restricted list is as follows: Restric_Control=π Passenger_ID,Restric_Reason (σ Passenger_ID∈Restric _ Data (Restric_Data)) Among them, Restric_Control represents the restricted list control information; π is a projection operation, indicating that a specific item is retained; Passenger_ID represents the passenger ID number; Restric_Reason represents the reason for being included in the restricted list; σ is a selection operation, used to filter data that meets the conditions; As long as the passenger is on the restricted list, control needs to be carried out, so the restricted list control information Restric_Control at this time is the final control information Final_Control; If the passenger is not on the restricted list, search in the special passenger data Special_Data; if the passenger is in the special passenger data after the search, the passenger is determined to be a controlled person and the control reason is obtained from the special passenger data. The control reason can be that the passenger is a VIP passenger or a dishonest passenger with bad credit. The control information obtained at this time is the final control information, and then the final control information is retained in the passenger data table to be verified. The formula for obtaining the control information is as follows: If(Restric_Control==null){Final_Control=π Passenger_ID,Special_Reason (σ Passenger_ID∈Special_Data (Special_Reason))} Among them, Restric_Control==null means that the restricted list control information is empty, that is, the passenger is not on the restricted list; Special_Reason indicates the reason for being listed as a special passenger.

[0038] S2. Passenger security verification is performed at the front end of the security information system, including the following steps: S2.1、Verify the identity information of passengers: collect the head portrait information in the passenger's ID information, compare it with the passenger's face information collected by the real-time camera, and verify the consistency between the passenger and the ID; Figure 2 As shown, the specific steps include: S2.1.1. Reading ID information: When a passenger passes through the security check counter, the passenger's ID information is read by a scanning device to extract the passenger's ID number and ID portrait information; the scanning device may be an ID card reader or an electronic boarding pass scanning device, and the passenger's ID may be a physical ID or an electronic ID. If it is a physical ID, the ID number and ID portrait information in the ID are directly read; if it is an electronic ID, the corresponding ID number and ID portrait information are extracted from the electronic document.

[0039] S2.1.2. Extract ID photo features: Use image processing algorithms to extract facial feature points from the ID photo information to form a digitized facial feature vector (e.g., a 128-dimensional or 512-dimensional feature vector), which is used in the subsequent feature comparison process.

[0040] S2.1.3. Face recognition and extraction: The camera installed at the security checkpoint collects the passenger’s current face image in real time, and uses image processing algorithms to extract the passenger’s facial features from the face image collected by the camera to generate a real-time face feature vector.

[0041] S2.1.4, Feature comparison: The facial feature vector extracted from the ID card is compared with the facial feature vector collected in real time by the camera, and the similarity between the two is calculated using the Euclidean distance or cosine similarity. The calculation formula for the Euclidean distance is:

[0042] Among them, f ID represents the facial feature vector extracted from the document, f face Represents the facial feature vector captured in real time; i is the feature number (index). When comparing features, there will be multiple features to be compared. For example, a person's facial features can be decomposed into several points or values, and i is the number of these features, such as the first feature, the second feature, etc. Represents the vector value of the face image of the ID card on the i-th feature; Represents the vector value of the passenger’s face photographed on site on the i-th feature; Euclidean distance can be simply understood as the "straight-line distance" between two points. In security verification, we can regard the ID photo and the real-time passenger photo as two points and calculate the "distance" between them. If this distance is very small, it means that the two are very similar.

[0043] The calculation formula for cosine similarity is:

[0044] Cosine similarity is used to compare the similarity between the face image on the ID and the face image captured by the camera. Its value is between -1 and 1, with 1 indicating complete similarity and 0 indicating complete difference. The closer to 1, the more similar. A similarity threshold can be set. If the calculated cosine similarity is higher than this similarity threshold, it is considered that the ID head portrait information matches the passenger's current facial information, and the verification is passed, allowing the next verification step; otherwise, the verification fails, and the security information system prompts the operator to conduct a manual review to further verify the passenger's identity.

[0045] S2.2. Verify the basic information of the passenger: Extract the passenger's itinerary information from the boarding pass provided by the passenger. The passenger's itinerary information includes the passenger's ID number, flight number, flight date, departure station, and terminal. Figure 3As shown, with the passenger's itinerary information as the index condition, the verification status information matching the passenger is obtained from the passenger data table to be verified for security verification, and the verification status information includes whether the passenger has checked in, whether the flight status is normal, whether the checked luggage has been opened, and whether control is being performed. Therefore, when checking whether the flight status is abnormal, if the security information system obtains that the passenger's flight status is abnormal, the security operator can be prompted that "the flight status is abnormal", and then the security operator prompts the passenger, or the security information system directly prompts the passenger that "the flight status is abnormal"; when checking whether the checked luggage is opened, if the security information system obtains that the passenger's checked luggage needs to be opened, the security operator is prompted to open the bag for inspection; when checking whether the passenger is a controlled person, if the security information system obtains that the passenger is a controlled person, the security operator is prompted to perform control.

[0046] Through all the above fusion algorithms and steps, the security inspection efficiency can be effectively improved.

[0047] The above description is only a preferred embodiment of the present invention and is not a limitation on the design of this case. Any equivalent changes made based on the design key of this case shall fall within the protection scope of this case.

Claims

1. A civil aviation passenger security verification method based on real-time fusion of multi-source heterogeneous data, characterized by: The following steps are involved: S1. At the back end of the security information system, data from different source systems are integrated to form a passenger data table to be verified. Specifically, the following steps are included: S1.

1. Input data: Each source system pushes original data to the security information system, including passenger booking data, passenger check-in data, flight schedule data, flight dynamic data, baggage security data, baggage transportation data, special passenger data, and restricted list passenger data; S1.2, Data preprocessing: Convert the data from different source systems into different formats, unify the data coding, field names and timestamps; S1.3, basic information integration: using data connection and projection, passenger booking data and check-in data are connected through passenger ID number, and the required basic passenger information is filtered into the passenger data table to be verified. The basic passenger information includes passenger ID number, flight number, flight date, departure station, terminal, and check-in status; S1.4, Flight data fusion: Merge flight dynamic data and flight plan data, select the latest flight status information to the passenger data table to be verified, and the flight status information can show whether the flight is normal or abnormal; S1.5, Baggage data fusion: Merge the baggage status information in the baggage security inspection data and the baggage transportation data, obtain the status information of whether the checked baggage needs to be unpacked and put it into the passenger data table to be verified; S1.6, Control data fusion: Search the restricted list passenger data and special passenger data by the passenger ID number to determine whether control is needed, and obtain the control information to the passenger data table to be verified. The control information includes whether control is needed and the reason for control; S2. Passenger security verification is performed at the front end of the security information system, including the following steps: S2.

1. Verify the passenger's identity information: collect the passenger's ID photo information and compare it with the passenger's face information collected by the real-time camera to verify the consistency between the passenger and the ID; S2.

2. Verify the basic information of the passenger: extract the passenger's itinerary information from the boarding pass provided by the passenger, the passenger's itinerary information includes the passenger's ID number, flight number, flight date, departure station, and terminal. Using the passenger's itinerary information as the index condition, obtain the passenger's verification status information from the passenger data table to be verified for security verification. The verification status information includes whether the passenger has checked in, whether the flight status is normal, whether the checked baggage has been unpacked, and whether control has been performed, so as to prompt the security operator whether to unpack and perform control operations.

2. A civil aviation passenger security verification method based on real-time fusion of multi-source heterogeneous data as claimed in claim 1, characterized in that: In step S1.3, the formula for generating the passenger basic information Passenger_Info is as follows: Among them, π is a projection operation, which means retaining specific items; σ is a selection operation, which is used to filter data that meets the conditions; Passenger_ID, Flight_Number, Flight_Date, Departure_Station, Depature_Terminal, and Is_Checked represent passenger ID number, flight number, flight date, departure station, terminal, and check-in status, respectively; It is a natural join operator, used for natural join operation; Seat_Data is passenger reservation data, and CheckIn_Data is check-in data.

3. The civil aviation passenger security verification method based on real-time fusion of multi-source heterogeneous data as claimed in claim 1, characterized in that: In step S1.4, the formula for obtaining the flight status information Flight_Status is: Flight_Status=π is_normal (σ latest (Flight_Dynamic ∪ Flight_Scheduled) ) Among them, π is_normal To retain flight status information and determine whether the flight status is normal; latest Indicates selecting the latest data based on timestamp; Flight_Dynamic is flight dynamic data; Flight_Scheduled is flight schedule data; ∪ indicates a merge operation.

4. A civil aviation passenger security verification method based on real-time fusion of multi-source heterogeneous data as claimed in claim 3, characterized in that: In step S1.4, if the filtered flight status information is planned, check-in started, boarding started, or arrived, the flight is in a normal state; if the filtered flight status information is canceled, boarding is closed, or taken off, the flight is in an abnormal state.

5. The civil aviation passenger security verification method based on real-time fusion of multi-source heterogeneous data as claimed in claim 1, characterized in that: In step S1.5, the baggage status information after combining the baggage security inspection data and the baggage transportation data includes two status information: abnormal baggage and no abnormal baggage. If the baggage status information contains information about abnormal baggage, it is determined that the baggage needs to be unpacked. If the baggage status information does not contain information about abnormal baggage, it is determined that the baggage does not need to be unpacked. The formula for obtaining the status information Baggage_Status of whether the checked baggage needs to be unpacked is: Baggage_Status=π Passenger_ID,Baggage_Check_Required (σ Baggage_Abnormal=True (Baggage_Data)∪σ Baggage_Abnormal=False (Baggage_Data)) Among them, π Passenger_ID,Baggage_Check_Required It indicates the operation of judging whether the baggage is unpacked and retaining the information of whether the baggage is unpacked and the passenger's ID number; σ is a selection operation, which is used to filter the data that meets the conditions; Baggage_Abnormal=True indicates that the baggage has an abnormality; Baggage_Abnormal=False indicates that the baggage has no abnormality; Baggage_Data indicates the baggage status information after merging the baggage security inspection data and the baggage transportation data; ∪ indicates a merge operation.

6. A civil aviation passenger security verification method based on real-time fusion of multi-source heterogeneous data as claimed in claim 5, characterized in that: In step S1.5, if the luggage security inspection data records information about prohibited items, oversized items, or overweight, the luggage is abnormal and needs to be unpacked; if the luggage transportation data records information about damaged luggage, the luggage is abnormal and needs to be unpacked.

7. The civil aviation passenger security verification method based on real-time fusion of multi-source heterogeneous data as claimed in claim 1, characterized in that: In step S1.6, firstly, according to the passenger ID number Passenger_ID, the passenger data Restric_Data on the restricted list is retrieved to see whether the passenger is on the restricted list. If yes, the passenger is determined to be a controlled person and the control reason is obtained from the passenger data on the restricted list. Then, the control information is retained in the passenger data table to be verified. The formula for obtaining the restricted list control information is as follows: Restric_Control=π Passenger_ID,Restric_Reason (σ Passenger_ID∈Restric _ Data (Restric_Data)) Among them, Restric_Control represents the restricted list control information; π is a projection operation, indicating that a specific item is retained; Passenger_ID represents the passenger ID number; Restric_Reason represents the reason for being included in the restricted list; σ is a selection operation, used to filter data that meets the conditions; At this time, the restricted list control information Restric_Control is the final control information Final_Control; If the passenger is not in the restricted list, the special passenger data Special_Data is searched. If the passenger is in the special passenger data, the passenger is determined to be a controlled person and the control reason is obtained from the special passenger data to generate the final control information. The final control information is then retained in the passenger data table to be verified. The formula for obtaining the final control information is as follows: If(Restric_Control==null){Final_Control=π Passenger_ID,Special_Reason (σ Passenger_ID∈Special_Data (Special_Reason))} Among them, Restric_Control==null means that the restricted list control information is empty; Special_Reason indicates the reason for being listed as a special passenger; If the passenger is not included in the special passenger data, it is determined that the passenger is not a monitored person and does not need to be monitored.

8. The civil aviation passenger security verification method based on real-time fusion of multi-source heterogeneous data as claimed in claim 1, characterized in that: In step S2.1, verifying the passenger's identity information specifically includes the following steps: S2.1.1、Reading ID information: When a passenger passes through the security check counter, the ID information of the passenger is read through the scanning device to extract the ID number and ID photo information of the passenger; S2.1.

2. Extract ID photo features: Use image processing algorithms to extract facial feature points from the ID photo information to form a digital facial feature vector; S2.1.3, Face recognition extraction: The camera installed at the security checkpoint collects the passenger's current face image in real time, and uses the image processing algorithm to extract the passenger's facial features from the face image collected by the camera to generate a face feature vector; S2.1.4, Feature comparison: The facial feature vector extracted from the ID card is compared with the facial feature vector collected in real time by the camera, and the similarity between the two is calculated using the Euclidean distance or cosine similarity. The calculation formula for the Euclidean distance is: Among them, f ID represents the facial feature vector extracted from the document, f face Represents the face feature vector captured in real time; i is the sequence number of the feature; Represents the vector value of the ID’s facial image on the i-th feature; Represents the vector value of the passenger’s face photographed on site on the i-th feature; The calculation formula for cosine similarity is: A similarity threshold is set. If the calculated cosine similarity is higher than the similarity threshold, it is considered that the ID photo information matches the passenger's current facial information, the verification is passed, and the next verification step is allowed; otherwise, the verification fails, and the security information system prompts the operator to perform manual review to further verify the passenger's identity.

9. The civil aviation passenger security verification method based on real-time fusion of multi-source heterogeneous data as claimed in claim 1, characterized in that: In step S1.1, the source systems include a ticket booking system, a departure system, an airport information integration system, a baggage security inspection system, a baggage sorting system, a Civil Aviation Administration system and a public security system; the passenger reservation data originates from the ticket booking system, the passenger check-in information originates from the departure system, the flight schedule data and flight dynamic data originate from the airport information integration system, the baggage security inspection data originates from the baggage security inspection system, the baggage transportation data originates from the baggage sorting system, the special passenger data originates from the Civil Aviation Administration system, and the restricted list passenger data originates from the public security system.

Images