Method and device for desensitizing multi-modal data and personal privacy data of intelligent connected automobile
By adopting format-retaining encryption algorithms and dynamic key management in intelligent connected vehicles, the efficiency and security problems of traditional encryption technology in multimodal data processing are solved, and efficient and secure data sharing and privacy protection are achieved.
Patent Information
- Application Number
- CN202510818102.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-18
- Publication Date
- 2025-08-22
AI Technical Summary
The existing traditional encryption technology cannot flexibly process multimodal data in intelligent connected vehicles, resulting in inefficient encryption, excessive consumption of computing resources, and lack of effective hierarchical and dynamic management of key management, which increases data security risks.
The format reserved encryption (FPE) algorithm is used to classify and encrypt multimodal data, dynamically generate encryption keys, and implement hierarchical key management, combining permission control and privacy protection effectiveness evaluation to ensure the security and availability of data during the sharing process.
It improves the encryption efficiency of multimodal data, reduces computing resource consumption, enhances the security of keys, prevents key leakage and abuse, and improves the privacy protection effect during data sharing.
Smart Images

Figure CN120528679A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of automobile information security technology, and more specifically to a method and device for desensitizing multimodal personal privacy data in intelligent connected vehicles, which is suitable for protecting the security of multimodal personal privacy data in sharing and transmission scenarios. Background Art
[0002] With the rapid development of intelligent connected vehicle (ICV) technology, multimodal data generated by vehicles (such as audio, images, video, and text) is increasingly interacting between onboard systems and external platforms. This data contains a wealth of personal privacy information, such as the driver's voice commands, dashcam footage, vehicle location information, and the user's personal identity information. The transmission, storage, and sharing of this multimodal personal privacy data presents numerous security risks, including unauthorized access, theft by malicious attackers, and data leakage. Therefore, effectively ensuring the security of multimodal privacy data during sharing and transmission has become a key research direction in the field of ICV data security.
[0003] Existing traditional data encryption technologies, such as symmetric and asymmetric encryption, generally provide a certain degree of security, but these technologies face various challenges. For one thing, traditional encryption algorithms do not differentiate between data types and cannot flexibly process different types of data. For example, different data types such as audio, video, and text vary significantly in structure and format. Using the same encryption method may result in inefficient encryption or the inability to restore the data to its original format after decryption. Furthermore, traditional encryption technologies are complex, especially when involving large-scale data sharing and frequent real-time encryption and decryption operations in vehicle environments. This can lead to excessive consumption of computing resources, impacting system performance and user experience.
[0004] Furthermore, as the demand for data sharing among connected vehicles increases, the complexity of these scenarios is also increasing. Connected vehicles not only need to exchange data with cloud servers, other vehicles, and transportation infrastructure, but may also involve multi-party data sharing across regions and platforms. Therefore, ensuring data security requires more than just encryption; key management and updates must also be considered. Existing key management mechanisms often lack effective hierarchical and dynamic management when dealing with multi-party sharing, making them susceptible to key leakage, misuse, and unauthorized access, further increasing data security risks.
[0005] Therefore, there is an urgent need for an innovative data encryption and desensitization method that can not only ensure data security but also meet the real-time processing requirements of multimodal data of intelligent connected vehicles, especially in data sharing and transmission scenarios, to ensure the efficiency of the system and the privacy protection effect. Summary of the Invention
[0006] The present invention proposes a method, device, and storage medium for desensitizing personal privacy data in multimodal data of intelligent connected vehicles, which can solve at least one of the technical problems in the background technology.
[0007] To achieve the above object, the present invention adopts the following technical solutions:
[0008] A method for desensitizing personal privacy data of multimodal data of intelligent connected vehicles, comprising the following steps:
[0009] S1: Collect multimodal personal privacy data (text, images, videos, etc.) from intelligent connected vehicles and classify them according to data type and sensitivity so that appropriate encryption strategies can be adopted for different data types;
[0010] S2: Based on the FPE (Format Preserving Encryption) algorithm, a simplified encryption strategy is formulated for various types of data. The data is encrypted and its original format is preserved to ensure the operability of the encrypted data during use.
[0011] S3: Dynamically generate encryption keys based on different data sharing scenarios and implement hierarchical key management to ensure key security. Also, during the data sharing process, securely distribute keys to ensure secure access for all parties.
[0012] S4: After data sharing, the recipient performs decryption based on permissions to ensure secure data access. The privacy protection effectiveness of the entire encryption and sharing process is continuously evaluated and optimized to ensure efficient and flexible privacy protection.
[0013] Furthermore, the specific steps for determining the corresponding encryption strategy according to different data types are as follows:
[0014] S111: Perform preliminary analysis and type identification on the collected multimodal data, including images (such as driver's facial images, license plate images, etc.), videos (such as driver's facial videos, dashcam license plate videos, etc.), and text (such as vehicle latitude and longitude data, etc.), so as to formulate encryption schemes later.
[0015] S112: Sub-classify the data based on a preset classification algorithm and mark the information involving personal privacy to determine the specific data scope that needs to be protected.
[0016] Furthermore, the specific steps of formulating a simplified encryption strategy for various types of data based on the FPE (Format Preserving Encryption) algorithm in S2 are as follows:
[0017] S211: Select a simplified FPE algorithm suitable for intelligent connected vehicles and reduce the number of calculation steps by optimizing the algorithm structure.
[0018] S212: Set corresponding encryption rounds and key lengths for different data types (such as images and videos) to improve encryption efficiency while maintaining the data format.
[0019] Furthermore, the method for dynamically generating encryption keys and implementing hierarchical key management in S3 according to different data sharing scenarios includes the following steps:
[0020] S311: Divide the key hierarchy according to the data sensitivity level to ensure that the keys of highly sensitive data have higher security.
[0021] S312: Utilize a dynamic key generation mechanism to generate keys through a random number generator, and combine public key encryption and private key decryption to securely distribute the keys during the sharing process.
[0022] This invention desensitizes the multimodal personal privacy data of intelligent connected vehicles by following a series of steps: format-preserving encryption strategy development, dynamic key generation and hierarchical management, data sharing and permission control, and privacy protection effectiveness evaluation. This ensures data security during sharing. This solution facilitates seamless integration of multimodal data within in-vehicle systems, maximizing format compatibility and operational continuity for personal privacy data.
[0023] On the other hand, the present invention further discloses a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above method.
[0024] In summary, to address the above issues, the present invention proposes a method for desensitizing multimodal personal privacy data in intelligent connected vehicles. This method, based on multimodal data classification and a simplified format-preserving encryption algorithm, effectively protects the security and availability of personal privacy data while preserving the data's original format. The technical implementation scheme of the present invention is to construct an efficient, secure, and flexible data desensitization solution through steps such as multimodal data collection and classification, multi-data type encryption processing based on simplified format-preserving encryption, dynamic key generation and hierarchical key management, secure data sharing and permission control, and dynamic evaluation of privacy protection effectiveness.
[0025] The beneficial effects of the present invention are:
[0026] This invention classifies and encrypts multimodal data (such as audio, images, and video) in connected vehicles. Using a format-preserving encryption (FPE) algorithm, it ensures that each data type retains its original format after encryption, thereby ensuring data integrity and availability. Compared to traditional encryption technologies, this invention offers flexible processing based on different data types, significantly improving the security of encrypted multimodal data while maintaining system functionality.
[0027] This invention optimizes encryption and decryption efficiency and reduces unnecessary computational steps, ensuring efficient and responsive encryption operations, particularly in resource-constrained vehicle systems. Compared to traditional, complex encryption algorithms, this invention reduces computational resource consumption during the encryption process, making it suitable for the real-time encryption of large-scale data in intelligent connected vehicles and significantly improving system performance.
[0028] This invention utilizes a hierarchical key management and dynamic update mechanism to ensure key security during multi-party data sharing. Different data types use different key levels based on their sensitivity, and key updates are automatically triggered at different stages of data sharing, effectively preventing the risk of key leakage and misuse. Compared to existing fixed key management solutions, this invention significantly enhances key security through dynamic key generation and update, improving the overall protection level during data sharing.
[0029] This invention not only protects the privacy of multimodal data through FPE encryption but also ensures data security during sharing and transmission through key distribution and secure transmission protocols (such as TLS / SSL). Data remains encrypted during transmission, and only authorized recipients can decrypt and access the data, preventing man-in-the-middle attacks and unauthorized access. Furthermore, permission control and key usage monitoring mechanisms further prevent the risk of data misuse and enhance privacy protection throughout the entire data sharing process. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] Figure 1 It is a schematic diagram of the process of the present invention;
[0031] Figure 2 This is a framework diagram of the multimodal personal privacy data classification and encryption strategy for intelligent connected vehicles of the present invention;
[0032] Figure 3 This is a diagram of the dynamic key generation and distribution architecture of the present invention. DETAILED DESCRIPTION
[0033] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.
[0034] like Figure 1 As shown, the method for desensitizing personal privacy data of multimodal data of intelligent connected vehicles described in this embodiment performs the following steps through a computer device:
[0035] like Figure 1A method for desensitizing multimodal personal privacy data of an intelligent connected vehicle is shown, comprising the following steps:
[0036] S1: Collects data from multiple data sources in the vehicle system (such as onboard cameras and GPS sensors). Data types include: image data (driver's facial images, license plate images), video data (dashcam videos), and text location data (latitude and longitude information). Data is categorized based on data type and sensitivity, allowing for appropriate encryption strategies and priorities to be adopted for different data types. The sensitivity classification is shown in the following table.
[0037] Table 1 Data sensitivity classification
[0038]
[0039]
[0040] Design specific sensitivity scoring functions for different types of data:
[0041] S(d)=ω1·I(d)+ω2·C(d)+ω3·T(d)
[0042] Where I(d) represents the intrinsic sensitivity of the data, C(d) represents the context factor, T(d) represents the time sensitivity, and ω1, ω2, and ω3 are weight coefficients, satisfying ω1+ω2+ω3=1.
[0043] S2: Based on the FPE (Format Preserving Encryption) algorithm, a simplified encryption strategy is developed for various types of data, and the data is encrypted to ensure that the encrypted data can continue to be used in intelligent connected vehicles and the data format remains unchanged;
[0044] S3: Dynamically generate encryption keys based on different data sharing scenarios and implement hierarchical key management to ensure key security. Also, during the data sharing process, securely distribute keys to ensure secure access for all parties.
[0045] S4: After data sharing, the recipient performs decryption based on permissions to ensure secure data access. The privacy protection effectiveness of the entire encryption and sharing process is continuously evaluated and optimized to ensure efficient and flexible privacy protection.
[0046] Based on the needs of desensitizing and securely sharing multimodal personal privacy data of intelligent connected vehicles, the embodiments of the present invention propose a multimodal personal privacy data desensitization and encryption protection scheme with dynamic key management, aiming to meet the security and privacy requirements of different types of data in the intelligent connected vehicle environment. Taking into account the diversity and dynamism of intelligent connected vehicle data, the present invention uses a classified encryption strategy to perform format-preserving encryption (FPE) for structured and unstructured data respectively, ensuring that the data format is not changed during transmission and sharing. The embodiment first uses a data acquisition and classification module to classify the multimodal data generated by the vehicle according to the data type, selects the corresponding encryption strategy based on different data types (such as text, image, video, etc.), and realizes efficient encryption through a simplified FPE algorithm. In terms of key management, dynamic key generation and distribution technology is adopted to ensure that each encrypted data fragment has an independent encryption key, and through hierarchical key management and dynamic update mechanism, the key security is maintained during the data sharing process to prevent unauthorized access. Furthermore, to address the security and real-time requirements of various data sharing scenarios within intelligent connected vehicles (e.g., vehicle-to-cloud, vehicle-to-road), key distribution and usage are based on permission control, ensuring that legitimate users can decrypt and access shared data securely. Finally, a differential privacy protection evaluation mechanism is incorporated to provide optimized feedback on the desensitization and encryption of multimodal data, continuously improving privacy protection and enabling efficient and reliable secure sharing of multimodal data within the intelligent connected vehicle environment.
[0047] like Figure 2 As shown in FIG, a framework diagram of the multimodal personal privacy data classification and encryption strategy for intelligent connected vehicles according to an embodiment of the present invention is provided, and a detailed description is provided as follows:
[0048] In a specific embodiment, before performing S1 (collecting multimodal personal privacy data from smart connected vehicles, including text, images, videos, etc., and classifying them according to data type and sensitivity, so as to adopt appropriate encryption strategies for different data types), it is necessary to first pre-process the multimodal data generated by the input smart connected vehicle. The main purpose of this preprocessing stage is to ensure that the format of all input data is unified, and to eliminate potential redundant or noisy data, so as to improve the accuracy of subsequent processing. The data classification module divides the input multimodal personal privacy data into three categories: text data, image data, and video data according to the different characteristics of the data. The purpose of this classification is to formulate targeted encryption strategies according to different data types to ensure that efficient encryption protection can be achieved without affecting the data usage experience.
[0049] After classifying the data, sensitivity evaluation is performed according to the formula S(d) = ω1·I(d) + ω2·C(d) + ω3·T(d). Intrinsic sensitivity I(d) evaluates the amount of sensitive information contained in the data itself, and the calculation formula is:
[0050]
[0051] Where n is the number of sensitive information types in the data, ω i is the weight of the i-th category of sensitive information, v i is the integrity score of the i-th category of sensitive information (0 to 1), p i is the probability of existence of the i-th category of sensitive information (0 to 1).
[0052] Take the text data "Li Ming, phone number 13812345678, Zhangjiang Hi-Tech Park, Pudong New Area, Shanghai" as an example. This text contains three types of sensitive information:
[0053] Name (ω1=0.5), fully visible v1=1, sensitive information exists p1=1;
[0054] Phone (ω2=0.7), fully visible v2=1, sensitive information exists p2=1;
[0055] Position (ω3 = 0.8), fully visible v3 = 1, sensitive information exists p3 = 1;
[0056] So the inner sensitivity This means that the text data has an intrinsic sensitivity of 0.67, which is a relatively high level.
[0057] The context factor C(d) takes into account the impact of data usage scenarios and environments on sensitivity, and is calculated as follows:
[0058] C(d)=α·L(d)+β·S(d)+γ·N(d)
[0059] Among them, L(d) is the location sensitivity (0 to 1), S(d) is the scene sensitivity (0 to 1), N(d) is the network environment sensitivity (0 to 1), α, β, γ are weight coefficients, satisfying α+β+γ=1.
[0060] Assume that the data is collected in the workplace, daily driving scenarios, and using a mobile data network. Take weights α = 0.4, β = 0.2, γ = 0.2, L(d) = 0.7, S(d) = 0.5, N(d) = 0.6, so C(d) = 0.6.
[0061] Time sensitivity T(d) evaluates the sensitivity change of data over time and is calculated as:
[0062]
[0063] Among them, T baseis the base time sensitivity (usually 0.8-1.0), Δt is the time since the data was generated (in days), is the time decay coefficient, T real is the real-time sensitivity increment, f is the data access frequency (within 24 hours), and μ is the frequency impact coefficient.
[0064] Assume that the data is user location information, generated 6 hours ago (Δt = 0.25 days), and accessed 8 times in the past 24 hours (f = 8), T base =0.8, T real =0.3, μ=0.1, so T(d)=0.9592, indicating that the data has high time sensitivity due to its recent generation time and frequent access.
[0065] Therefore, according to the formula S(d) = ω1·I(d) + ω2·C(d) + ω3·T(d), with weights ω1 = 0.5, ω2 = 0.3, and ω3 = 0.2, the sensitivity of the data is S(d) = 0.67×0.5+0.6×0.3+0.9592×0.2=0.70684. According to the classification standards, the data is classified as "sensitive" and requires a higher level of desensitization protection, with a higher encryption priority.
[0066] In the encryption strategy selection module, different encryption strategies are selected for each data type. For example, for text data, the system uses a format-preserving encryption algorithm based on a combination of a balanced Feistel network structure and an unbalanced Feistel network structure for encryption. This algorithm can ensure that the encrypted data still retains the original text format, and ensures that the encrypted data can be directly used for further processing when decrypted at the end without the need for additional decoding steps. For image and video data, image recognition technology and video frame analysis are used to accurately locate sensitive areas. For sensitive information in the image (such as license plates, driver's faces, etc.), encryption is achieved through conversion of grayscale values or color values to ensure that the overall format of the image remains unchanged. In the encryption process of video data, the system first identifies key frames (such as license plates or frames containing people) through frame analysis, and then encrypts these key frames and keeps the format of the video unchanged to ensure the continuity and consistency of the video stream. The details are as follows:
[0067] For text data, text data is usually structured, such as telephone numbers, ID numbers, bank account numbers, etc. Take the text data X = "13812345678" as an example. First, block processing is performed to divide the data into three parts: X = (x1, x2, x3) = ("138", "1234", "5678"), based on the permutation function F kAnd the key K = 1234, generate the encrypted data Y: y i =F k (x i ), first encrypt x1 = "138", block length n = 3, so 10 n =1000, we get y1=(x1+K)mod1000=(138+1234)mod1000=1372mod1000="372". Similarly, x2="1234" is encrypted, and the block length n=4, so 10 n =10000, we get
[0068] y2=(x2+K)mod10000=(1234+1234)mod10000=2468mod10000="2468", x3="5678" is encrypted, the block length n=4, so 10 n =10000, we get
[0069] y3=(x3+K)mod10000=(5678+1234)mod10000=6912mod10000="6912", the three encrypted parts are concatenated in order to form the final encryption result Y="37224686912".
[0070] For image data, the image data is stored in the form of pixel matrix. Encryption needs to ensure that the image dimension remains unchanged while protecting sensitive information (such as face, license plate). Assume that the original image matrix The secret key K=50 is to encrypt the sensitive area in the lower right corner (2*2 area) Keeping other areas unchanged, use the permutation function F of the additive mapping k (P ij )=(P ij +K)mod256,P′ 22 =(200+50)mod256=250,P′ 23 =(180+50)mod256=230, P′ 32 =(150+50)mod256=200, P′ 33 =(75+50)mod256=125, the encrypted image matrix is
[0071] For video data, the video data consists of continuous frames, and format-preserving encryption must ensure that the encrypted video frame format and playback smoothness remain unchanged. The original video consists of several frames F t Each frame is a 4*4 grayscale image. Here, a certain frame For example, the secret key K=3, the goal is to encrypt the sensitive area in the lower right corner (2*2 area) Keeping other areas unchanged, use the permutation function F of the multiplication mapping k (P ij )=(P ij *K)mod256,P′ 33 =(110*3)mod256=330mod256=74, P′ 34 =(120*3)mod256=360mod256=104, P′ 43 =(150*3)mod256=450mod256=194, P′ 44 =(160*3)mod256=480mod256=224, encrypted video frame matrix
[0072] Through this architecture diagram, you can clearly see how the system selects the optimal encryption strategy based on the data type, ensuring that data is effectively protected during sharing and transmission, and can quickly restore the original information after decryption, achieving efficient and secure data management.
[0073] like Figure 3 As shown in the figure, the dynamic key generation and distribution architecture diagram of the embodiment of the present invention is described in detail as follows:
[0074] In a specific embodiment, the dynamic key generation and distribution architecture consists of a key generation module, a key storage and management module, a key update and destruction module, and a key distribution module. These modules work together to build an efficient and secure key management system.
[0075] First, the key generation module is responsible for generating keys for different purposes, including global keys, group keys, and personal keys. Global keys apply to all nodes in the system, ensuring basic communication security. Group keys are used for data exchange within a specific group, improving the efficiency and security of intra-group communication. Personal keys are used for independent encryption operations for users or devices, ensuring the privacy of personalized data. Key generation uses a high-strength random number generation algorithm to ensure unpredictability and uniqueness.
[0076] The generated keys are then centrally managed by the Key Storage and Management Module. Specifically, the keys are first stored in an encrypted key vault, where they are protected by high-strength encryption to minimize the risk of key leakage. Furthermore, a permission control mechanism ensures that different users and devices can only access keys within their authorized scope, ensuring secure and compliant key usage.
[0077] To address key leakage risks or periodic key renewal needs, the Key Renewal and Destruction Module plays a key role. This module consists of a Key Renewal Trigger and a Key Destruction Module. The Key Renewal Trigger initiates the key renewal process based on pre-set rules (such as time periods, usage counts, or anomaly detection), regenerating new keys from the Key Generation Module. The Key Destruction Module ensures that old keys are completely destroyed upon expiration to prevent malicious exploitation.
[0078] When keys need to be distributed, the key distribution module is responsible for securely transmitting them to relevant devices or nodes (such as vehicles, roadside equipment, and the cloud). This key distribution process utilizes public key encryption and private key decryption mechanisms, distributed over secure communication channels (such as TLS / SSL protocols) to ensure that keys are not intercepted or tampered with during transmission. Furthermore, authentication and permission control mechanisms are implemented to ensure that only authorized recipients can access the keys.
[0079] In summary, this architecture builds a highly secure and flexible key management system through a full-process design of dynamic key generation, distribution, update, and destruction, which can effectively meet the encryption and sharing needs of multimodal data in intelligent connected vehicle scenarios.
[0080] In another aspect, the present invention further discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the above method.
[0081] On the other hand, the present invention further discloses a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above method.
[0082] In another embodiment provided in the present application, a computer program product comprising instructions is also provided, which, when executed on a computer, enables the computer to execute any of the methods for desensitizing personal privacy data of multimodal data of intelligent connected vehicles in the above embodiments.
[0083] It is understandable that the system, device and storage medium provided in the embodiments of the present invention correspond to the method provided in the embodiments of the present invention, and the explanation, examples and beneficial effects of the relevant contents can refer to the corresponding parts of the above methods.
[0084] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When software is used for implementation, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website, computer, server or data center to another website, computer, server or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrations. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).
[0085] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.
[0086] Each embodiment in this specification is described in a related manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the system embodiment is generally similar to the method embodiment, so the description is relatively simple. For related parts, refer to the description of the method embodiment.
[0087] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.
Claims
1. A method for desensitizing personal privacy data in multimodal data of intelligent connected vehicles, characterized in that: The following steps are included: S1: Collect multimodal personal privacy data from intelligent connected cars and classify them according to data type and sensitivity; S2: Based on the FPE algorithm, a simplified encryption strategy is developed for various types of data. The data is encrypted and maintained in its original format to ensure operability after encryption. S3: Dynamically generates encryption keys based on different data sharing scenarios and implements hierarchical key management to ensure key security. Furthermore, during the data sharing process, keys are securely distributed to ensure secure access for all parties. S4: After data sharing, the recipient performs decryption operations based on permissions to ensure secure data access, and continuously evaluates and optimizes the privacy protection effect throughout the entire encryption and sharing process to ensure the efficiency and flexibility of privacy protection.
2. The method for desensitizing personal privacy data in multimodal data of intelligent connected vehicles according to claim 1, characterized in that: Step S1 specifically includes: S111: Conduct preliminary analysis and type identification of collected multimodal data, including images, videos, and text, to facilitate the subsequent development of encryption solutions; S112: Sub-classify the data based on a preset classification algorithm and mark the information involving personal privacy to determine the specific data scope that needs to be protected.
3. The method for desensitizing personal privacy data in multimodal data of intelligent connected vehicles according to claim 2, characterized in that: Step S2 specifically includes: S211. Select an FPE algorithm suitable for intelligent connected vehicles and reduce the number of computational steps by optimizing the algorithm structure. S212: Set corresponding encryption rounds and key lengths for different data types to improve encryption efficiency while maintaining the data format.
4. The method for desensitizing personal privacy data in multimodal data of intelligent connected vehicles according to claim 3, characterized in that: Step S3 specifically includes: S311: Divide the key hierarchy according to the data sensitivity level to ensure that the keys of highly sensitive data have higher security; S312: Utilize a dynamic key generation mechanism to generate keys through a random number generator, and combine public key encryption and private key decryption to securely distribute the keys during the sharing process.
5. The method for desensitizing personal privacy data in multimodal data of intelligent connected vehicles according to claim 2, characterized in that: The image data includes a driver's facial image and a license plate image, the video data includes a driving recorder video, and the text location data includes latitude and longitude information.
6. The method for desensitizing personal privacy data in multimodal data of intelligent connected vehicles according to claim 1, characterized in that: It also includes designing sensitivity scoring functions for different types of data: S(d)=ω1·I(d)+ω2·C(d)+ω3·T(d) Where I(d) represents the intrinsic sensitivity of the data, C(d) represents the context factor, T(d) represents the time sensitivity, ω1, ω2, ω3 are weight coefficients, satisfying ω1+ω2+ω3=1; Intrinsic sensitivity I(d) evaluates the amount of sensitive information contained in the data itself and is calculated as: Where n is the number of sensitive information types in the data, ω i is the weight of the i-th category of sensitive information, v i is the integrity score of the i-th category of sensitive information (0 to 1), p i is the probability of existence of the i-th category of sensitive information.
7. The method for desensitizing personal privacy data in multimodal data of intelligent connected vehicles according to claim 1, characterized in that: The dynamic key generation and distribution in step S3 includes a key generation module, a key storage and management module, a key update and destruction module, and a key distribution module; First, the key generation module is responsible for generating keys for different purposes, including global keys, group keys and personal keys. The global key applies to all nodes in the system, ensuring the security of basic communications; The group key is used for data interaction within a specified group, improving the efficiency and security of intra-group communication; The personal key is used for independent encryption operations of users or devices to ensure the privacy of personalized data; The key is generated using a high-intensity random number generation algorithm to ensure the unpredictability and uniqueness of the key; Subsequently, the generated keys are centrally managed by the key storage and management module; The key update and destruction module consists of a key update trigger and a key destruction module; The key update trigger starts the key update process based on preset rules including time period, number of uses or anomaly detection, and regenerates new keys from the key generation module; The key destruction module ensures that old keys are completely destroyed after they expire to prevent them from being used maliciously; In scenarios where keys need to be distributed, the key distribution module is responsible for securely transmitting the keys to relevant devices or nodes.
8. The method for desensitizing personal privacy data in multimodal data of intelligent connected vehicles according to claim 7, characterized in that: The generated keys are centrally managed by the key storage and management module. Specifically, the keys are first stored in the encryption key library, and the keys themselves are protected by high-strength encryption technology to avoid the risk of key leakage. At the same time, combined with the permission control mechanism, different users and devices can only access keys within their authorized scope, thereby ensuring the security and compliance of key use.
9. The method for desensitizing personal privacy data in multimodal data of intelligent connected vehicles according to claim 7, characterized in that: Public key encryption and private key decryption mechanisms are used in the key distribution process, and key distribution is completed through a secure communication channel to ensure that the key is not intercepted or tampered with during transmission. In addition, identity authentication and permission control mechanisms are introduced to ensure that only legitimate recipients can obtain the key.
10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the computer program is executed by the processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 9.
Citation Information
Cited By
Target identification system suitable for vehicle image privacy compliance detection
CN121392799A
