Low-altitude intelligent aircraft data security protection method and system, medium and server

Abstract

The invention provides a data security protection method and system for a low-altitude intelligent aircraft, a medium and a server, and relates to the technical field of data security protection, and the method comprises the steps: analyzing internal data, judging whether the internal data contains image data, if yes, storing the image data, and if not, storing the image data; according to the method, the image data is stored in the stack mode, when the user identity verification is passed, the pixel value corresponding to the stack is fed back, and when the user identity verification is passed, the internal data is stored, and then when the user calls the data, the user identity is verified, and if the verification is passed, the related data is fed back, and otherwise, an alarm is executed. And the security of the image data is improved.

Description

Technical Field

[0001] The present application relates to the technical field of protecting data security, and in particular to methods, systems, media and servers for protecting data security of low-altitude intelligent aircraft. Background Art

[0002] With the rapid development of low-altitude intelligent aircraft technology, such aircraft have shown unprecedented potential for wide application in many fields such as aerial photography, agricultural monitoring, emergency rescue, cargo transportation and even leisure and entertainment, greatly improving work efficiency and quality of life. However, while fully enjoying the convenience and benefits brought by these high-tech products, an issue that cannot be ignored - data security, is gradually emerging and becoming one of the key factors restricting the further development of low-altitude intelligent aircraft.

[0003] Since low-altitude intelligent aircraft are often equipped with wide-angle high-definition cameras, sensors and other advanced equipment when performing missions, they can capture and transmit a large amount of ground images, videos and other sensitive information in real time. If this information is improperly used or leaked, it may pose a serious threat to individuals, companies and even social stability. Especially in some special periods or sensitive areas, no-fly zones are set up to maintain public safety, protect important facilities or respond to specific events.

[0004] Although low-altitude intelligent aircraft will comply with regulations after receiving a no-fly instruction and no longer enter these clearly defined no-fly zones, even if the low-altitude intelligent aircraft itself does not cross the boundary, it is still possible for the low-altitude intelligent aircraft to capture content that is prohibited from being photographed in the edge area or even in the no-fly zone when approaching the no-fly boundary, allowing the low-altitude intelligent aircraft to capture clear images when far away from the target area. Once these images containing no-fly zones are illegally obtained, disseminated, or maliciously tampered with, the consequences will be disastrous. Summary of the invention

[0005] In order to improve the security of the internal data of low-altitude intelligent aircraft, the present application provides a low-altitude intelligent aircraft data security protection method, system, medium and server.

[0006] In the first aspect, the present application provides a method for protecting the data security of low-altitude intelligent aircraft, which adopts the following technical solutions: The low-altitude intelligent aircraft data security protection method includes the following steps: Data collection: obtain internal data stored by low-altitude intelligent aircraft; First judgment: judging whether the internal data contains image data, if so, executing the step of establishing the system; if not, executing the second step of storing; Establishing a coordinate system: Establishing a coordinate system with any pixel point in the image data as the origin of the coordinate system; First acquisition: acquiring the coordinate value and pixel value of each pixel point in the image data; Stack building: Build stacks equal to the number of pixel value categories; First storage: push the coordinate values ​​of the pixel points with the same pixel value into the stack, and record the pixel value corresponding to each stack; Second storage: store internal data in a preset database; User verification: Verify whether the user's identity is legal. If yes, execute the feedback step; if not, execute the alarm step; Feedback: Feedback the pixel value and / or internal data corresponding to each stack to the user; Alarm: Send an alarm signal.

[0007] The present application first obtains the internal data stored in the low-altitude intelligent aircraft, and then determines whether there is image data in the internal data. If so, a coordinate system is established to obtain the coordinate value and pixel value of each pixel point, and then a stack with the same number of categories as the pixel value is established, so as to conveniently store and manage the coordinate values ​​of pixel points with the same pixel value, which helps to quickly access and process these pixel points. The use of the stack also improves the flexibility and efficiency of image data processing. Subsequently, the present application pushes the coordinate values ​​​​that are the same as the pixel values ​​into the stack in a preset order, and records the pixel values ​​corresponding to each stack. Otherwise, the internal data is stored, and the correct data will be fed back to the user only after the user identity verification is passed. By adopting the above technical solution, the present application realizes the classified storage and management of pixel points, and then realizes the hiding of image data, which improves the security of internal image data of low-altitude intelligent aircraft.

[0008] Optionally, after executing the step of data collection and before executing the step of first judgment, the method further includes: Clustering: clustering the internal data using the K-Means clustering algorithm to obtain a category label for each internal data; First calculation: perform mean removal processing on the internal data under the same category label to obtain the internal data after the first processing; perform variance normalization processing on the internal data after the first processing to obtain the internal data after the second processing, which is recorded as the first data; Construct a matrix: Construct a covariance matrix based on the first data, wherein the pth row and qth column element of the covariance matrix The calculation formula is as follows: ; in, is the value of the pth first data under the k1th category label; is the mean of all first data under the k1th category label; is the value of the qth first data under the k2th category label; is the mean of all first data under the k2th category label; n is the number of category labels; Second calculation: perform eigenvalue decomposition on the covariance matrix to obtain the largest eigenvalue and the corresponding eigenvector; The third calculation: performing a product operation on the first data and the characteristic vector to obtain an operation result, and updating the operation result as internal data.

[0009] This application first performs a mean removal process on the internal data under the same category label to eliminate the mean shift in the internal data under the same category label, making the internal data more centralized and easy to analyze, and the variance normalization process further eliminates the dimensional differences of the internal data in different dimensions, making the internal data comparable in different dimensions, thereby improving the accuracy and stability of subsequent data analysis. Subsequently, this application constructs a covariance matrix based on the internal data after secondary processing to quantify the correlation and dependency between different secondary processed internal data, and then performs eigenvalue decomposition on the covariance matrix to obtain eigenvalues ​​and corresponding eigenvectors, and then multiplies the secondary processed internal data with the eigenvector, thereby mapping the secondary processed internal data to the space represented by the eigenvector. This application can hide the original internal data while retaining the main information in the internal data. Even if an illegal user obtains the first data after dimensionality reduction, the original dimension of the first data cannot be known, thereby further improving the security of the internal data.

[0010] Optionally, after executing the first acquisition step and before executing the stack building step, the method further includes: Fourth calculation: calculate the variance of all pixel values ​​in the image data; Second judgment: judging whether the variance is equal to zero, if so, executing the step of stacking; if not, executing the step of image segmentation; Image segmentation: Use region growing algorithm to segment image data and obtain multiple segmentation blocks; The third judgment: judging whether there is a noise point in the segmented block based on the eight-neighborhood pixel values, if so, modifying the pixel value at the noise point to the average value of the eight-neighborhood pixel values ​​of the noise point; if not, executing the image update step; Image update: Update each segmented block to the image data respectively and perform the fourth calculation step.

[0011] The present application first calculates the variance of all pixel values ​​in the image data. A variance of zero indicates that the pixel values ​​in the image data are the same or change uniformly, that is, the image is a monochrome image or a uniform image. Subsequently, the present application determines whether the variance is equal to zero to identify whether the current image is a monochrome image or a uniform image. When the variance is not zero, it indicates that the current image is an uneven image, that is, there may be noise points, and further image segmentation is required to determine whether there are indeed noise points. Subsequently, the present application uses a regional growth algorithm to segment the image data to obtain multiple segmentation blocks with similar pixel values, and then processes and analyzes different areas in the image separately. After that, it is determined whether there are noise points in the current segmentation block based on the eight-neighborhood pixel values. If so, the pixel value at the noise point is modified to the average value of the pixel values ​​of its eight-neighborhood pixel points, thereby smoothing the image. Through image segmentation and noise processing, the present application can reduce noise and interference in the image and improve the quality and availability of the image.

[0012] Optionally, after performing the step of image segmentation and before performing the step of third judgment, the method further includes: Fourth judgment: judge whether the number of pixel points in the segmented block is one, if so, execute the third judgment step; if not, execute the image update step.

[0013] The present application determines whether the number of pixels in the segmentation block is one. If so, it means that the segmentation block is a monochrome block, and directly executes the third judgment step to reduce the number of execution times of the image segmentation step, thereby improving the overall processing efficiency. The present application can optimize the allocation of computing resources by first screening out the segmentation blocks containing only one pixel, so that more resources can be concentrated on processing larger segmentation blocks containing more useful information, thereby improving processing efficiency.

[0014] Optionally, after executing the first judging step and before executing the second storing step, the method further includes: Second acquisition: obtain the clustering results of internal data, and obtain n cluster centers based on the clustering results; Database construction: construct n data sub-databases, and record the i-th data sub-database as the i-th data sub-database; Fifth calculation: sequentially calculate the Euclidean distances between the j-th cluster center and the remaining cluster centers to obtain n-1 Euclidean distances; calculate the sum of the n-1 Euclidean distances and record it as the second data; First sorting: arranging the n second data in a preset order, and recording the result of the arrangement as a second data sequence; Association: Associating the second data in the second data sequence with the data sub-database having the same sequence number to obtain an association result; Third storage: store the internal data into the corresponding data sub-database according to the association results.

[0015] The present application first obtains the clustering results of the internal data, and obtains n cluster centers based on the clustering results. Subsequently, the present application constructs n data sub-libraries to facilitate the classified storage of the internal data of the low-altitude intelligent aircraft. Subsequently, the present application calculates the Euclidean distance between each cluster center, and respectively calculates the sum of the Euclidean distances from each cluster center to the remaining cluster centers, recorded as the second data, and arranges the n second data in a preset order. Subsequently, the present application associates the second data in the second data sequence with the data sub-library with the same sequence number, realizes the correspondence between the internal data and the data sub-library, and facilitates subsequent internal data storage and access. Subsequently, the present application stores the internal data in the corresponding data sub-library according to the association results, realizes the classified storage of the data, and then helps to speed up the data retrieval speed and improve the data processing efficiency. The present application enhances the security of the data through a dynamic and fine-grained data storage method.

[0016] Optionally, after performing the third storage step, the method further includes: First encryption: using a symmetric encryption algorithm to encrypt the internal data stored in the i-th data sub-database to obtain the i-th encrypted data, and using the i-th encrypted data to update the internal data stored in the i-th data sub-database; Second encryption: using an asymmetric encryption algorithm to encrypt the i-th encrypted data and the internal data stored in the i+1-th data sub-database to obtain the i+1-th encrypted data, and using the i+1-th encrypted data to update the internal data stored in the i+1-th data sub-database; Fifth judgment: judge whether i+1 is equal to n, if not, execute the iterative step; Iteration: Use the i+1th data sub-database as the new i-th data sub-database, use the i+1th encrypted data as the new i-th encrypted data, and perform the second encryption step until the preset stop condition is met.

[0017] The present application adopts a symmetric encryption algorithm to encrypt the internal data stored in the i-th data sub-library, which can improve the security of the internal data stored in the i-th data sub-library, so that even if the encrypted internal data is illegally obtained, it is difficult to be decrypted and tampered, thereby protecting the privacy and integrity of the internal data. Subsequently, the present application adopts an asymmetric encryption algorithm to encrypt the i-th encrypted data and the internal data stored in the i+1-th data sub-library, further enhancing the security of the data. Using different encryption algorithms can increase the difficulty of data being cracked and improve the hierarchical nature of data protection. Subsequently, by executing the fifth judgment step, the present application can check whether all internal data have completed the encryption process. If there is still unencrypted internal data, the encryption process is continued through iterative steps until all internal data is encrypted, thereby reducing the risk of data leakage. By adopting the above scheme, the present application further enhances the security of the internal data of low-altitude intelligent aircraft. The multi-level and multi-algorithm encryption method adopted by the present application improves the strength of data protection, making it difficult to decrypt and tamper with the internal data even if it is illegally obtained.

[0018] Optionally, after executing the associating step and before executing the first storing step, the method further includes: Data screening: Identify sensitive and non-sensitive data in internal data; Second sorting: sort the sensitive data in each cluster according to the order of collection time to obtain n sensitive data sequences; Calculate the distance: Calculate the distance matrix between the ath sensitive data sequence and the remaining sensitive data sequences respectively, and calculate the cumulative distance matrix based on the distance matrix; Determine the sequence: Based on the cumulative distance matrix, use the dynamic path planning algorithm to obtain the sensitive data sequence with the greatest similarity to the a-th sensitive data sequence, which is recorded as the similar data sequence; Combining: combining the sensitive data of the a-th sensitive data sequence with the non-sensitive data corresponding to the similar data sequence to obtain combined data, and updating the combined data as internal data.

[0019] The present application first identifies the sensitive data and non-sensitive data in the internal data, and then sorts the sensitive data in each cluster in chronological order to reveal the timing characteristics and changing trends of the data, which is convenient for subsequent data analysis and mining. Subsequently, the present application calculates the distance matrix between the a-th sensitive data sequence and the remaining sensitive data sequences, and calculates the cumulative distance matrix based on the distance matrix, thereby quantifying the similarity between the a-th sensitive data sequence and the remaining sensitive data sequences. Subsequently, the present application obtains the sensitive data sequence (i.e., similar data sequence) with the greatest similarity to the a-th sensitive data sequence based on the cumulative distance matrix, and then combines the sensitive data of the a-th sensitive data sequence with the non-sensitive data corresponding to the similar data sequence, and updates it to internal data, which helps to better utilize the non-sensitive data of other data sequences to hide the sensitive data of the a-th data sequence while protecting the sensitive data, thereby further improving data security.

[0020] In the second aspect, the present application provides a low-altitude intelligent aircraft data security protection system, which adopts the following technical solutions: Low-altitude intelligent aircraft data security protection system, including: Data acquisition module, used to obtain internal data stored in low-altitude intelligent aircraft; A first judging module is used to judge whether the internal data contains image data, and if so, trigger the system building module; otherwise, trigger the second storage module; A system building module is used to build a coordinate system using any pixel point in the image data as the coordinate origin; A first acquisition module is used to acquire the coordinate value and pixel value of each pixel point in the image data; A stacking module, used to build stacks with the same number of categories as pixel values; A first storage module, used for pushing the coordinate values ​​of pixel points with the same pixel value into a stack, and recording the pixel value corresponding to each stack; A second storage module, used for storing internal data in a preset database; The user verification module is used to verify whether the user's identity is legal. If it is legal, the feedback module is triggered; otherwise, the alarm module is triggered; A feedback module, used for feeding back the pixel value and / or internal data corresponding to each stack to the user; The alarm module is used to send out an alarm signal.

[0021] The present application realizes the classified storage and processing of the internal data of the aircraft through a series of processing of the image data, thereby improving the security of the internal image data of the aircraft.

[0022] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program is used to implement the described method when processed by a processor.

[0023] In a fourth aspect, the present application provides a server, on which the system is loaded, and the server includes: a processor, and a memory communicatively connected to the processor; The computer-readable storage medium is provided in the memory, and the computer-readable storage medium stores a computer program; When the processor processes the computer program stored on the computer-readable storage medium, the method described above is implemented.

[0024] In summary, the present application includes at least one of the following beneficial technical effects: 1. This application first obtains the internal data stored in the low-altitude intelligent aircraft, and then determines whether there is image data in the internal data. If so, a coordinate system is established to obtain the coordinate value and pixel value of each pixel point, and then a stack with the same number of categories as the pixel value is established, so as to conveniently store and manage the coordinate values ​​of pixel points with the same pixel value, which helps to quickly access and process these pixel points. The use of stacks also improves the flexibility and efficiency of image data processing.

[0025] 2. This application pushes the coordinate values ​​that are the same as the pixel values ​​into the stack in a preset order, and records the pixel values ​​corresponding to each stack. Otherwise, the internal data is stored. The correct data will be fed back to the user only after the user identity verification is passed. By adopting the above technical solution, this application realizes the classified storage and management of pixel points, and then realizes the hiding of image data, thereby improving the security of internal image data of low-altitude intelligent aircraft. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a flow chart of Example 1 of the present application; Figure 2 is a flow chart of Example 2 of the present application; Figure 3 is a flow chart of Example 3 of the present application; Figure 4 is a flow chart of Example 4 of the present application; Figure 5 This is a flow chart of Example 5 of the present application. DETAILED DESCRIPTION

[0027] The following combination Figures 1 to 5 This application is described in further detail.

[0028] Embodiment 1: This embodiment discloses a method for protecting data security of low-altitude intelligent aircraft, referring to Figure 1 , the method includes: S1 data collection, S2 first judgment, S3 image storage, S4 second storage and S5 legal verification, first collect internal data, and then judge whether the internal data contains image data. If it contains image data, select any pixel point in the image as the origin to build a system, obtain the coordinates and pixel values ​​of each pixel, and establish a stack equal to the number of pixel value categories to store the coordinate information of the same pixel value; if it does not contain image data, directly store the internal data in a preset database. In addition, this embodiment also includes a user identity verification link, legal users can obtain data feedback, and illegal users trigger an alarm signal. This embodiment includes the following steps: S1 data acquisition, obtaining stored internal data from the aircraft, the internal data including flight parameters, sensor readings, equipment status information, and collected image data, etc.

[0029] S2 is a first determination, determining whether the internal data contains image data, if so, executing S3 image storage; otherwise, executing S4 second storage.

[0030] S3 image storage includes S31 system building, S32 first acquisition, S33 stack building and S34 first storage.

[0031] S31 establishes a system, selecting an arbitrary pixel point in the image data as a coordinate origin. The coordinate origin can be any pixel point in the upper left corner, center or other position of the image.

[0032] Determine the direction of the coordinate axis. In this embodiment, the horizontal direction is selected as the horizontal axis and the vertical direction is selected as the vertical axis to form a rectangular coordinate system.

[0033] In a rectangular coordinate system, each pixel has a unique coordinate value that is used to identify its position in the image data.

[0034] In other embodiments, other directions may be selected as the horizontal axis and the vertical axis, as long as it can be determined that each pixel in the image data has a unique coordinate value.

[0035] S32 first obtains, traverses each pixel point in the image data, and obtains the coordinate value and pixel value of each pixel point.

[0036] S33 builds a stack, counting the number of different pixel values ​​in the image data, and for each different pixel value, builds a corresponding stack, the stack is used to store the coordinate values ​​of the pixel points with the same pixel value. Each element stored in the stack represents the position of the pixel point with the pixel value in the image.

[0037] S34 is the first storage. For each pixel point in the image data, the corresponding stack is found according to its pixel value, the coordinate values ​​of the pixel points with the same pixel value are pushed into the same stack, and the pixel value corresponding to each stack is recorded.

[0038] S4 Secondary storage stores internal data in a preset database.

[0039] S5 legal verification, including S51 user verification, S52 feedback and S53 alarm.

[0040] S51 user verification, verifying whether the user identity is legal, if yes, execute S52 feedback; if not, execute S53 alarm.

[0041] The user identity legitimacy verification can be achieved by entering a user name and password, using biometrics or other identity verification methods, and the standard for the user identity legitimacy is that the user successfully passes the verification.

[0042] S52 feedback: feeding back the pixel value and / or internal data corresponding to each stack to the user.

[0043] S53 alarm, sends out an alarm signal.

[0044] This embodiment implements the function of securely storing internal data stored in a low-altitude intelligent aircraft (hereinafter referred to as an "aircraft"), strengthens the management and protection of the aircraft's internal data, and improves the security of the internal data.

[0045] Example 2: Reference Figure 2 The difference between this embodiment and the first embodiment is that after executing S1 data collection and before executing S2 first judgment, it also includes: S61 clustering, after obtaining the internal data, the K-Means clustering algorithm is used to process these internal data. The K-Means clustering algorithm is an unsupervised learning algorithm used to divide data points into n clusters (the center of each cluster becomes a cluster center). After clustering, the category label of each internal data can be obtained.

[0046] S62 is the first calculation, which performs a mean removal process on the internal data under the same category label to obtain the internal data after the first processing; performs a variance normalization process on the internal data after the first processing to obtain the internal data after the second processing, which is recorded as the first data.

[0047] De-meaning refers to subtracting the mean of all internal data under the category from each internal data to eliminate the offset of the internal data. Variance normalization refers to dividing the internal data after de-meaning by the standard deviation of the internal data to scale the internal data to a uniform scale.

[0048] S63 constructs a matrix, constructing a covariance matrix based on the first data, wherein the p-th row and q-th column element of the covariance matrix The calculation formula is as follows: ; in, is the value of the pth first data under the k1th category label; is the mean of all first data under the k1th category label; is the value of the qth first data under the k2th category label; is the mean of all first data under the k2th category label; n is the number of category labels.

[0049] S64 is a second calculation, performing eigenvalue decomposition on the covariance matrix to obtain the maximum eigenvalue and the corresponding eigenvector.

[0050] S65 is a third calculation, performing a product operation on the first data and the feature vector to obtain an operation result, and updating the operation result as internal data.

[0051] This embodiment obtains the label of each internal data and performs dimensionality reduction processing on the data under different category labels. While reducing the number of processes, each internal data is disguised, thereby improving the security of the internal data.

[0052] Example 3: Reference Figure 3 The difference between this embodiment and embodiment 1 is that after executing S32 for the first acquisition and before executing S33 for stack building, the following is further included: S71: fourth calculation, calculating the variance of all pixel values ​​in the image data.

[0053] S72 is the second judgment, judging whether the variance is equal to zero. If so, it indicates that the current image is a monochrome image or a uniform image, and then S33 stacking is executed; if not, it indicates that the current image is a non-uniform image, and then S73 image segmentation is executed.

[0054] S73 image segmentation uses a region growing algorithm to segment the image data to obtain multiple segmentation blocks.

[0055] S74 is the fourth judgment, judging whether the number of pixel points in the segmentation block is one. If so, it indicates that the current segmentation block has been segmented into the minimum segmentation block and cannot be segmented further. Then, S75 is executed as the third judgment, judging whether there are noise points in the current segmentation block. If not, it indicates that the current segmentation block can continue to be segmented, and then S76 is executed to update the image.

[0056] S75 is the third judgment, judging whether there is a noise point in the segmented block based on the eight-neighborhood pixel values ​​of the current segmented block. If so, modify the pixel value at the noise point to the average value of the eight-neighborhood pixel values; if not, execute S76 image update.

[0057] S76 image update, respectively update each segmented block to image data, and execute S71 fourth calculation, until all segmented blocks are updated and then execute S33 stack building.

[0058] This embodiment first calculates the variance of all pixel values ​​in the image data, and determines whether the variance is zero to determine whether the image is a monochrome image or a uniform image. If so, a stacking operation is performed, otherwise image segmentation is performed; the image segmentation uses a region growing algorithm to obtain multiple segmentation blocks, and then the number of pixel points in the segmentation block is determined. If it is one, it indicates that it has been segmented into the smallest block and whether there is a noise point. If there is, the pixel value of the noise point is modified to the average value of the eight neighborhoods. Otherwise, the image data is updated and the steps of calculating the variance are repeated until all segmentation blocks are updated and stacking is performed. Through image segmentation and noise processing, this embodiment can reduce noise and interference in the image and improve the quality and usability of the image.

[0059] Example 4: Reference Figure 4 The difference between this embodiment and embodiment 2 is that after executing S2 first judgment and before executing S4 second storage, it also includes: S81 second acquisition, obtaining the clustering result of the internal data in the clustering in S61, and obtaining n cluster centers based on the clustering result.

[0060] S82 builds a database. According to the n cluster centers obtained by K-Means clustering, n data sub-databases are constructed. Each data sub-database is used to store internal data associated with the corresponding cluster center. The i-th data sub-database is recorded as the i-th data sub-database (i=1, 2,..., n).

[0061] S83 The fifth calculation is to calculate, for each cluster center, the Euclidean distance between it and the remaining cluster centers, and obtain n-1 Euclidean distances (because each cluster center has a distance from the other n-1 distance centers).

[0062] The values ​​of these Euclidean distances are added together to obtain a sum, which is recorded as the second data. This process is repeated for each cluster center to eventually obtain n second data.

[0063] S84 first sorting, arranging the obtained n second data in a preset order. The preset order is set based on actual needs, such as sorting from small to large or from large to small, and the sorted result is called a second data sequence.

[0064] S85: Associating each second data in the second data sequence with a data sub-library having the same sequence number to obtain an association result. For example, the second data with a sequence number of one in the second data sequence is associated with the first data sub-library, the second data with a sequence number of two is associated with the second data sub-library, and so on, until each data sub-library is associated with one second data.

[0065] S86: third storage. Since each data sub-library is associated with a specified second data, the second data corresponds to a specified cluster center, and the cluster center corresponds to a specified internal data, the internal data can be stored in the corresponding data sub-library through the association result.

[0066] The present embodiment is described below with reference to a specific case.

[0067] Assume that a low-altitude intelligent aircraft (hereinafter referred to as the aircraft) records a large amount of flight data, including parameters such as flight altitude, speed, temperature, pressure, as well as the readings of various sensors and equipment status information.

[0068] S1 data collection,collects 1,000 internal data records from the aircraft, each record contains flight parameters (such as,altitude, speed), sensor readings (such as temperature, pressure), and equipment status information (such as,engine status).

[0069] These internal data were processed by S61 clustering using the K-Means clustering algorithm (n=5 was set, i.e. the internal data were divided into 5 clusters), and 5 cluster centers were obtained. Each cluster center represents the average value of a group of internal data with similar characteristics.

[0070] S81 second acquisition, obtains the clustering results of the internal data, and obtains 5 cluster centers.

[0071] S82 database construction, based on the 5 cluster centers obtained by K-Means clustering, 5 corresponding data sub-databases were constructed.

[0072] Each data sub-database will be used to store data points associated with the corresponding cluster center. For the convenience of management and reference, the i-th data sub-database is recorded as the i-th data sub-database (i=1, 2, 3, 4, 5).

[0073] S83 first calculation, for each cluster center (for example, the first cluster center), calculate the Euclidean distance between it and the remaining four cluster centers to obtain four Euclidean distance values, then add up the four Euclidean distance values ​​to obtain a sum, and record the sum as the second data.

[0074] This step calculates all five cluster centers and finally obtains five second data.

[0075] S84 is a first sorting step, in which the five obtained second data are arranged in ascending order to obtain an ordered data sequence, namely, a second data sequence.

[0076] S85: Associating each second data in the second data sequence with a data sub-library having the same sequence number. For example, the second data sequence is: altitude data, speed data, temperature data, pressure data, and engine status data; the second data corresponding to the altitude data is associated with the first data sub-library, the second data corresponding to the speed data is associated with the second data sub-library, the second data corresponding to the temperature data is associated with the third data sub-library, the second data corresponding to the pressure data is associated with the fourth data sub-library, and the second data corresponding to the engine status data is associated with the fifth data sub-library.

[0077] S86 third storage, according to the association result, the original internal data is stored in the corresponding data sub-library. For example, the height data is stored in the first data sub-library, the speed data is stored in the second data sub-library, the temperature data is stored in the third data sub-library, the pressure data is stored in the fourth data sub-library, and the engine status data is stored in the fifth data sub-library.

[0078] This embodiment involves obtaining internal data (such as flight parameters, sensor readings, equipment status information, etc.) from an aircraft, using the K-Means clustering algorithm to process these internal data, obtaining n cluster centers, and constructing n data sub-libraries, each of which is associated with a corresponding cluster center. Then, the Euclidean distance between each cluster center and other cluster centers is calculated, and the sum of these Euclidean distances is calculated to obtain n second data, and then the second data is sorted to form a second data sequence. The position of the second data in the second data sequence is associated with the sequence number of the data sub-library, and finally the internal data is stored in the corresponding data sub-library according to the association result, so as to realize dynamic storage and efficient management of data, thereby improving the security of the internal data of the low-altitude intelligent aircraft.

[0079] Example 5: Reference Figure 5 The difference between this embodiment and embodiment 4 is that after executing the step of associating in S85 and before executing the third storage in S86, the method further includes: S91 data screening conducts in-depth analysis of the internal data from the aircraft to identify sensitive and non-sensitive data.

[0080] The identification method for sensitive data can be an identification method based on data characteristics and usage, an identification method based on preset rules, or an identification method based on a machine learning model.

[0081] For example, the identification process based on data characteristics and usage is as follows: First, the internal data of the aircraft is classified in detail, such as flight parameters, personnel information, system status data, etc. Then, based on the characteristics and usage of the data, a preliminary judgment is made as to which data may be sensitive data and which data may be non-sensitive data. Then, the specific content of the data is analyzed, such as whether it contains personal privacy information (such as name, ID number, contact information, etc.), aircraft operation data (such as flight plan, etc.) or aircraft safety data (such as fault records, maintenance records, etc.). Finally, according to the sensitivity of the data, it is divided into sensitive data and non-sensitive data.

[0082] For example, the identification process based on the machine learning algorithm is: using existing sensitive data and non-sensitive data samples to train a supervised learning model. The trained supervised learning model is used to identify the internal data to determine whether it contains sensitive data.

[0083] For example, the rule-based recognition process is: First, formulate sensitive data identification rules: Based on the characteristics and uses of the data inside the aircraft, formulate a set of sensitive data identification rules, which may include requirements for the content, format, storage location, etc. of the data.

[0084] Then apply identification rules: Apply identification rules to the internal data of the aircraft to identify sensitive data through matching and filtering.

[0085] S92 second sorting: within each cluster, the sensitive data is sorted in the order of collection time to obtain n sensitive data sequences, each sensitive data sequence corresponds to a cluster center.

[0086] S93 calculates the distance. For the ath sensitive data sequence, this step will calculate the distance matrix between it and the remaining sensitive data sequences. Each element in the distance matrix represents a certain distance measurement (such as Euclidean distance, Manhattan distance, etc.) between the corresponding samples of two sensitive data sequences. The corresponding samples can be samples with the same sequence number in each sensitive data sequence, or samples with the same collection time in each sensitive data sequence.

[0087] Based on a single distance matrix, this step further calculates a cumulative distance matrix, which reflects the overall similarity or difference between sensitive data sequences.

[0088] S94 determines the sequence. According to the cumulative distance matrix, this step can find the sensitive data sequence with the greatest similarity to the a-th sensitive data sequence through a dynamic path planning algorithm, which is recorded as a similar data sequence.

[0089] S95 combines the sensitive data of the a-th sensitive data sequence with the non-sensitive data corresponding to the similar data sequence to obtain the disguised a-th data sequence. The disguised a-th sensitive data sequence will be updated to internal data, and then S86 third storage is executed to store these internal data into the corresponding data sub-library.

[0090] S96: first encryption, using a symmetric encryption algorithm to encrypt the internal data stored in the i-th data sub-database, obtaining the i-th encrypted data and the i-th encrypted key, and using the i-th encrypted data to update the internal data stored in the i-th data sub-database.

[0091] S97 second encryption, based on the first encryption in S96, uses an asymmetric encryption algorithm to encrypt the i-th encrypted data and the internal data stored in the i+1-th data sub-library, obtains the i+1-th encrypted data and the i+1-th encrypted key, and uses the i+1-th encrypted data to update the internal data stored in the i+1-th data sub-library.

[0092] S98 is the fifth judgment, judging whether i+1 is equal to n. If so, no subsequent processing is performed; if not, it means that there is still unencrypted internal data, and S99 iteration needs to be executed to continue encryption.

[0093] In S99 iteration, the i+1th data sub-database is used as the new ith data sub-database, and the first encryption is performed in S96. This process will continue until a preset stop condition is met (such as all data sub-databases have been encrypted, the maximum number of iterations has been reached, etc.).

[0094] When a user calls for relevant data, the user's identity will be verified first, and then the decryption will be performed in reverse iteration starting from the last data sub-database using the corresponding decryption algorithm and key until all data sub-databases are decrypted. Subsequently, the data will be reorganized and verified to identify and protect sensitive data.

[0095] This embodiment first identifies the sensitive data in the internal data, then combines the sensitive data in the current data sequence with the non-sensitive data in the similar data sequence to obtain a disguised current data sequence, then updates the disguised current data sequence to the data sub-library where the current data sequence is stored, and then uses iterative encryption to store all disguised current data sequences to improve the security of internal data.

[0096] Embodiment 6: This embodiment discloses a low-altitude intelligent aircraft data security protection system, the system comprising: The data acquisition module is responsible for collecting and storing internal data from the aircraft. These data include but are not limited to flight parameters (such as speed, altitude, heading, etc.), sensor data (such as temperature, pressure, humidity, etc.), system logs, fault diagnosis information, etc.

[0097] The first judgment module is used to judge whether the internal data contains image data according to a preset rule, and if so, trigger the system building module; otherwise, trigger the second storage module.

[0098] The system building module is used to build a coordinate system using any pixel point in the image data as the coordinate origin.

[0099] The first acquisition module is used to acquire the coordinate value and pixel value of each pixel point in the image data.

[0100] The stacking module is used to build stacks with the same number as the categories of pixel values, and each stack is responsible for storing one category of pixel values.

[0101] The first storage module is used to push the coordinate values ​​of pixel points with the same pixel value into a stack, and record the pixel value corresponding to each stack.

[0102] The second storage module is used to store the internal data in a preset database.

[0103] The user verification module is used to verify whether the user's identity is legal. If it is legal, the feedback module is triggered; otherwise, the alarm module is triggered.

[0104] The feedback module is used to feed back the pixel value and / or internal data corresponding to each stack to the user.

[0105] The alarm module is used to send out an alarm signal.

[0106] This embodiment analyzes the internal data to determine whether the internal data contains image data. If it contains image data, the image data is stored, otherwise the internal data is stored. Then, when the user calls the data, the user identity is verified. If the verification is successful, the relevant data is fed back, otherwise an alarm is executed. This application uses a stack to store image data. Only when the user identity verification is successful, the pixel value corresponding to the stack will be fed back, thereby improving the security of the image data.

[0107] Embodiment 7: This embodiment provides a computer-readable storage medium on which a computer program is stored. When the computer program is processed by a processor, it is used to implement the described method.

[0108] Embodiment 8: This embodiment provides a server, on which the system is loaded, and the server includes: a processor, and a memory communicatively connected to the processor; The computer-readable storage medium is provided in the memory, and the computer-readable storage medium stores a computer program; When the processor processes the computer program stored on the computer-readable storage medium, the method described above is implemented.

[0109] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for protecting data security of low-altitude intelligent aircraft, characterized in that: include: Data collection: obtain internal data stored by low-altitude intelligent aircraft; First judgment: judging whether the internal data contains image data, if so, executing the step of establishing the system; if not, executing the second step of storing; Establishing a coordinate system: Establishing a coordinate system with any pixel point in the image data as the origin of the coordinate system; First acquisition: acquiring the coordinate value and pixel value of each pixel point in the image data; Stack building: Build stacks equal to the number of pixel value categories; First storage: push the coordinate values ​​of the pixel points with the same pixel value into the stack, and record the pixel value corresponding to each stack; Second storage: store internal data in a preset database; User verification: Verify whether the user's identity is legal. If yes, execute the feedback step; if not, execute the alarm step; Feedback: Feedback the pixel value and / or internal data corresponding to each stack to the user; Alarm: Send an alarm signal.

2. The low-altitude intelligent aircraft data security protection method according to claim 1 is characterized in that: After executing the step of data collection and before executing the step of first judgment, the method further includes: Clustering: clustering the internal data using the K-Means clustering algorithm to obtain a category label for each internal data; First calculation: perform mean removal processing on the internal data under the same category label to obtain the internal data after the first processing; perform variance normalization processing on the internal data after the first processing to obtain the internal data after the second processing, which is recorded as the first data; Construct a matrix: Construct a covariance matrix based on the first data, wherein the pth row and qth column element of the covariance matrix The calculation formula is as follows: ； in, is the value of the pth first data under the k1th category label; is the mean of all first data under the k1th category label; is the value of the qth first data under the k2th category label; is the mean of all first data under the k2th category label; n is the number of category labels; Second calculation: perform eigenvalue decomposition on the covariance matrix to obtain the largest eigenvalue and the corresponding eigenvector; The third calculation: performing a product operation on the first data and the characteristic vector to obtain an operation result, and updating the operation result as internal data.

3. The low-altitude intelligent aircraft data security protection method according to claim 2 is characterized in that: After executing the first acquisition step and before executing the stack building step, the method further includes: Fourth calculation: calculate the variance of all pixel values ​​in the image data; Second judgment: judging whether the variance is equal to zero, if so, executing the step of stacking; if not, executing the step of image segmentation; Image segmentation: Use the region growing algorithm to segment the image data and obtain multiple segments; The third judgment: judging whether there is a noise point in the segmented block based on the eight-neighborhood pixel values, if so, modifying the pixel value at the noise point to the average value of the eight-neighborhood pixel values ​​of the noise point; if not, executing the image update step; Image update: Update each segmented block to the image data respectively and perform the fourth calculation step.

4. The low-altitude intelligent aircraft data security protection method according to claim 3 is characterized in that: After performing the step of image segmentation and before performing the step of third determination, the method further includes: Fourth judgment: judge whether the number of pixel points in the segmented block is one, if so, execute the third judgment step; if not, execute the image update step.

5. The low-altitude intelligent aircraft data security protection method according to any one of claims 2 to 4, characterized in that: After executing the first judging step and before executing the second storing step, the method further includes: Second acquisition: obtain the clustering results of internal data, and obtain n cluster centers based on the clustering results; Database construction: construct n data sub-databases, and record the i-th data sub-database as the i-th data sub-database; Fifth calculation: sequentially calculate the Euclidean distances between the j-th cluster center and the remaining cluster centers to obtain n-1 Euclidean distances; calculate the sum of the n-1 Euclidean distances and record it as the second data; First sorting: arranging the n second data in a preset order, and recording the result of the arrangement as a second data sequence; Association: Associating the second data in the second data sequence with the data sub-database having the same sequence number to obtain an association result; Third storage: store the internal data into the corresponding data sub-database according to the association results.

6. The low-altitude intelligent aircraft data security protection method according to claim 5 is characterized in that: After executing the third storage step, the method further includes: First encryption: using a symmetric encryption algorithm to encrypt the internal data stored in the i-th data sub-database to obtain the i-th encrypted data, and using the i-th encrypted data to update the internal data stored in the i-th data sub-database; Second encryption: using an asymmetric encryption algorithm to encrypt the i-th encrypted data and the internal data stored in the i+1-th data sub-database to obtain the i+1-th encrypted data, and using the i+1-th encrypted data to update the internal data stored in the i+1-th data sub-database; Fifth judgment: judge whether i+1 is equal to n, if not, execute the iterative step; Iteration: Use the i+1th data sub-database as the new i-th data sub-database, use the i+1th encrypted data as the new i-th encrypted data, and perform the second encryption step until the preset stop condition is met.

7. The low-altitude intelligent aircraft data security protection method according to claim 6 is characterized in that: After executing the associating step and before executing the first storing step, the method further includes: Data screening: Identify sensitive and non-sensitive data in internal data; Second sorting: sort the sensitive data in each cluster according to the order of collection time to obtain n sensitive data sequences; Calculate the distance: Calculate the distance matrix between the ath sensitive data sequence and the remaining sensitive data sequences respectively, and calculate the cumulative distance matrix based on the distance matrix; Determine the sequence: Based on the cumulative distance matrix, use the dynamic path planning algorithm to obtain the sensitive data sequence with the greatest similarity to the a-th sensitive data sequence, which is recorded as the similar data sequence; Combining: combining the sensitive data of the a-th sensitive data sequence with the non-sensitive data corresponding to the similar data sequence to obtain combined data, and updating the combined data as internal data.

8. A low-altitude intelligent aircraft data security protection system, the system is used to implement the method according to any one of claims 1 to 7, characterized in that: include: Data acquisition module, used to obtain internal data stored in low-altitude intelligent aircraft; A first judgment module, used for judging whether the internal data contains image data, and triggering a system building module if the internal data contains image data; Otherwise, the second storage module is triggered; A system building module is used to build a coordinate system using any pixel point in the image data as the coordinate origin; A first acquisition module is used to acquire the coordinate value and pixel value of each pixel point in the image data; A stacking module, used to build stacks with the same number of categories as pixel values; A first storage module, used for pushing the coordinate values ​​of pixel points with the same pixel value into a stack, and recording the pixel value corresponding to each stack; A second storage module, used for storing internal data in a preset database; The user verification module is used to verify whether the user's identity is legal. If it is legal, the feedback module is triggered; Otherwise, the alarm module is triggered; A feedback module, used for feeding back the pixel value and / or internal data corresponding to each stack to the user; The alarm module is used to send out an alarm signal.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is processed by a processor, it is used to implement the method according to any one of claims 1 to 7.

10. A server, characterized in that: The server is loaded with the system according to claim 8, and the server comprises: a processor, and a memory communicatively connected to the processor; The memory is provided with a computer-readable storage medium according to claim 9, and the computer-readable storage medium stores a computer program; When the processor processes the computer program stored on the computer-readable storage medium, the method according to any one of claims 1 to 7 is implemented.

Images