Data leakage detection method and terminal for intelligent network connection automobile data
By extracting image feature components, restoring digital watermark images, and verifying evidence sets from the intelligent connected vehicle data platform, the privacy protection and efficiency issues in the detection of data leakage in intelligent connected vehicles are solved, and effective data security identification and cost control are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-30
- Publication Date
- 2026-03-31
AI Technical Summary
Existing technologies for detecting data leaks in intelligent connected vehicles suffer from problems such as the risk of original data leakage, low detection efficiency, high cost, and false leak reports. In particular, they are difficult to effectively identify data leaks in multi-party sharing environments.
By extracting the target feature components of the image to be identified from a public platform, requesting the data owner to provide watermark embedding data to restore the digital watermark image, and verifying the validity of the evidence set through a smart contract, it is determined whether the image is a leaked image.
Effectively identify data breaches, protect the privacy of transaction images, prevent false reports of breaches, improve data security, and incentivize users to report breaches through a reward mechanism, thereby reducing detection costs.
Smart Images
Figure CN121765772A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data processing technology, and in particular to a data leakage detection method and terminal for intelligent connected vehicle data. Background Technology
[0002] With the rapid development of vehicle-to-everything (V2X) technology and intelligent transportation systems, the sharing of intelligent connected vehicle data has become crucial for improving traffic management efficiency, enhancing the driving experience, and promoting the development of intelligent driving technology. However, the security and privacy protection of intelligent connected vehicle data under multi-party sharing are becoming increasingly prominent issues, especially data breaches. Without an effective accountability mechanism for data breaches, not only will user privacy be compromised, but it may also lead to a series of security problems such as data infringement and abuse, data tampering and forgery. Although there has been some research and practice on data breach accountability mechanisms in the context of multi-party sharing of intelligent connected vehicle data, many challenges and problems still exist.
[0003] Existing research methods require complete original data to assist in the extraction of evidence of data breaches, which poses a risk of original data leakage. In terms of leakage tracking, there are also various problems. For example, relying on copyright holders to track leaks independently results in low leakage detection efficiency, while outsourcing to monitoring platforms results in excessively high leakage detection costs. Crowdsourcing solutions, which offer a cost-effective compromise, have issues such as false reporting of leaks. Summary of the Invention
[0004] The technical problem to be solved by the present invention is to provide a data leakage detection method and terminal for intelligent connected vehicle data, which can effectively identify leaked data and improve data security while ensuring the privacy of transaction data.
[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:
[0006] Data leakage detection methods for intelligent connected vehicle data include:
[0007] The image to be identified is acquired from a public platform, and the target feature components of the image to be identified at different frequencies are extracted.
[0008] The image to be identified and the target feature component are sent to the data owner of the image to be identified in order to request the watermark embedding data corresponding to the target feature component in the original image corresponding to the image to be identified.
[0009] The digital watermark image of the image to be identified is reconstructed based on the watermark embedding data, and the target evidence set is extracted from the digital watermark image.
[0010] The validity of the target evidence set is verified by a smart contract. If the target evidence set is valid, the image to be identified is determined to be a leaked image.
[0011] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows:
[0012] A data leakage detection terminal for intelligent connected vehicle data includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements each step of the data leakage detection method for intelligent connected vehicle data described above.
[0013] The beneficial effects of this invention are as follows: After a suspected leaked image to be identified is detected on a public platform, target feature components of the image to be identified at different frequencies are extracted. Then, based on the target feature components, watermark embedding data is requested from the data owner to reconstruct the digital watermark image that may correspond to the image to be identified. After extracting the target evidence set from the digital watermark image, the validity of the target evidence set is verified through a smart contract. If the target evidence set is valid, it indicates that the image to be identified is indeed data leaked by the digital owner. Since this invention obtains watermark embedding data through target feature components, the data owner does not need to provide all the features of the transaction image during the process of assisting in the reconstruction of the digital watermark data. Instead, only some features of the transaction image are required. This avoids other data owners from reconstructing the transaction image based on all the features of the transaction image, effectively protecting the privacy of the transaction image. At the same time, based on some features of the transaction image, the digital watermark image of the image to be identified is simulated in advance, thereby extracting the target evidence set from the digital watermark image to determine whether the image to be identified is a leaked image. This avoids the problem of false reporting of leaks, achieves effective identification of leaked data, and improves data security. Attached Figure Description
[0014] Figure 1 A flowchart of a data leakage detection method for intelligent connected vehicle data provided in an embodiment of the present invention;
[0015] Figure 2 A schematic diagram of a transaction environment provided in an embodiment of the present invention;
[0016] Figure 3 This is a data interaction diagram of a data leakage detection method provided in an embodiment of the present invention;
[0017] Figure 4 This invention provides a schematic diagram of the structure of a data leakage detection terminal for intelligent connected vehicle data, as provided in an embodiment of the present invention.
[0018] Label Explanation:
[0019] 100. Data leakage detection terminal for intelligent connected vehicle data; 101. Memory; 102. Processor. Detailed Implementation
[0020] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.
[0021] Embodiments of the present invention provide a data leakage detection method for intelligent connected vehicle data, including:
[0022] The image to be identified is acquired from a public platform, and the target feature components of the image to be identified at different frequencies are extracted.
[0023] The image to be identified and the target feature component are sent to the data owner of the image to be identified in order to request the watermark embedding data corresponding to the target feature component in the original image corresponding to the image to be identified.
[0024] The digital watermark image of the image to be identified is reconstructed based on the watermark embedding data, and the target evidence set is extracted from the digital watermark image.
[0025] The validity of the target evidence set is verified by a smart contract. If the target evidence set is valid, the image to be identified is determined to be a leaked image.
[0026] As described above, the beneficial effects of this invention are as follows: After a public platform identifies a suspected leaked image to be identified, it extracts target feature components of the image at different frequencies, and then requests watermark embedding data from the data owner based on these target feature components. This watermark embedding data is then used to reconstruct the digital watermark image that may correspond to the image to be identified. After extracting the target evidence set from the digital watermark image, a smart contract verifies the validity of the evidence set. If the evidence set is valid, it indicates that the image to be identified is indeed data leaked by the digital owner. Because this invention obtains watermark embedding data through target feature components, the data owner does not need to provide all features of the transaction image during the process of assisting in the reconstruction of the digital watermark data. Instead, it only needs to provide some features, preventing other data owners from reconstructing the transaction image based on all its features, thus effectively protecting the privacy of the transaction image. Simultaneously, based on some features of the transaction image, the digital watermark image embedded with the digital watermark is pre-simulated, thereby extracting the target evidence set from the digital watermark image to determine whether the image to be identified is a leaked image. This avoids the problem of falsely reporting leaks, achieving effective identification of leaked data and improving data security.
[0027] Further, the extraction of target feature components at different frequencies of the image to be identified includes:
[0028] Perform discrete wavelet transform on the image to be identified to obtain the low-frequency feature components and high-frequency feature components of the image to be identified;
[0029] Select low-frequency target feature components and high-frequency target feature components from the low-frequency feature components and the high-frequency feature components.
[0030] As described above, performing discrete wavelet transform on the image to be identified facilitates the extraction of feature components of different frequencies in the image. At the same time, based on the target feature components of different frequencies, the digital watermark image can be accurately reconstructed, ensuring the effective identification of leaked data.
[0031] Furthermore, the watermark embedding data includes low-frequency original feature components in the original image corresponding to the low-frequency target feature components, high-frequency original feature components in the original image corresponding to the high-frequency target feature components, and embedding weight parameters in the original image corresponding to the low-frequency feature components and the high-frequency feature components when embedding digital watermarks.
[0032] The step of restoring the digital watermark image of the image to be identified based on the watermark embedding data includes:
[0033] The low-frequency watermark feature components of the image to be identified are generated based on the low-frequency target feature components, the low-frequency original feature components, and the embedding weight parameters corresponding to the low-frequency feature components.
[0034] The high-frequency watermark feature components of the image to be identified are generated based on the high-frequency target feature components, the high-frequency original feature components, and the embedding weight parameters corresponding to the high-frequency feature components.
[0035] The digital watermark image of the image to be identified is reconstructed based on the low-frequency watermark feature components and the high-frequency watermark feature components.
[0036] As described above, the data owner does not need to disclose all feature components of the original transaction image during the process of assisting in the reconstruction of the digital watermark image. This prevents other data owners from reconstructing the original transaction image using the image data provided by the data owner, thus ensuring the privacy of the original transaction image. Furthermore, the digital watermark image can be simulated based on the partial watermark embedding data provided by the data owner, and then verified using the evidence set corresponding to the digital watermark image. This avoids the problem of false reporting and leakage, and achieves effective identification of leaked data.
[0037] Further, the step of reconstructing the digital watermark image of the image to be identified based on the low-frequency watermark feature components and the high-frequency watermark feature components includes:
[0038] The low-frequency watermark feature components and the high-frequency watermark feature components are subjected to inverse wavelet transform to obtain the digital watermark image of the image to be identified.
[0039] As described above, the digital watermark image of the image to be identified is simulated and restored by inverse wavelet transform of the corresponding feature components. The data restoration method is simple and ensures detection efficiency.
[0040] Furthermore, the extraction of the target evidence set from the digital watermark image includes:
[0041] Calculate the average pixel value of the digital watermark image, and generate a target evidence set based on the pixels in the digital watermark image whose pixel values are greater than the average pixel value.
[0042] As described above, the target evidence set is extracted by the average pixel value of the digital watermark image. This ensures the privacy of the digital watermark image while facilitating subsequent verification of whether the image to be identified is leaked data.
[0043] Furthermore, verifying the validity of the target evidence set via a smart contract includes:
[0044] The standard evidence set pre-stored in the original image is obtained through a smart contract, and the intersection between the standard evidence set and the target evidence set is calculated.
[0045] If the intersection is less than a preset standard value, the target evidence set is determined to be invalid.
[0046] If the intersection is not less than the standard value, then the target evidence set is determined to be valid.
[0047] As described above, the similarity between the standard evidence set and the target evidence set is determined by calculating the intersection. A high similarity indicates the target evidence set is valid, while a low similarity indicates it is invalid. Verifying the validity of the target evidence set using the standard evidence set avoids the problem of other data sources falsely reporting leaked data, effectively improving the accuracy of leaked data detection.
[0048] Furthermore, after determining that the target evidence set is valid, the process also includes:
[0049] Mark the standard evidence set corresponding to the target evidence set;
[0050] Before calculating the intersection between the standard evidence set and the target evidence set, the method further includes:
[0051] Detect whether the marker exists in the standard evidence set;
[0052] If it exists, exit the verification and end the data breach detection;
[0053] If it does not exist, then perform the step of calculating the intersection between the standard evidence set and the target evidence set.
[0054] As described above, before a smart contract verifies the target evidence set, it is necessary to confirm whether the standard evidence set has been effectively presented. If the standard evidence set has been marked, it indicates that the image corresponding to the standard evidence set has been leaked, that is, the standard evidence set has been effectively presented. Therefore, the verification is terminated, the data leakage detection is ended, and the same image to be identified is not repeatedly verified.
[0055] Furthermore, if the target evidence set is valid, it also includes:
[0056] In response to the payment instruction of the smart contract, the system receives the digital currency corresponding to the payment instruction and sends the digital currency to the target account that uploaded the image to be authenticated.
[0057] As described above, for target accounts that provide valid evidence, rewards are given in the form of digital currency, thereby incentivizing more users to obtain suspected leaked data on public platforms and increasing the breadth of the scope of leaked data monitoring.
[0058] Furthermore, it also includes:
[0059] If the target evidence set is invalid, a deduction instruction is sent to the target account to request the deduction of a preset amount of digital currency from the target account.
[0060] As described above, for target accounts with invalid evidence, a certain amount of digital currency needs to be deducted to cover the costs of data breach detection, thereby reducing the cost of data breach detection.
[0061] Another embodiment of the present invention provides a data leakage detection terminal for intelligent connected vehicle data, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the various steps in the above-described data leakage detection method for intelligent connected vehicle data.
[0062] As described above, the beneficial effects of this invention are as follows: After a public platform identifies a suspected leaked image to be identified, it extracts target feature components of the image at different frequencies, and then requests watermark embedding data from the data owner based on these target feature components. This watermark embedding data is then used to reconstruct the digital watermark image that may correspond to the image to be identified. After extracting the target evidence set from the digital watermark image, a smart contract verifies the validity of the evidence set. If the evidence set is valid, it indicates that the image to be identified is indeed data leaked by the digital owner. Because this invention obtains watermark embedding data through target feature components, the data owner does not need to provide all features of the transaction image during the process of assisting in the reconstruction of the digital watermark data. Instead, it only needs to provide some features, preventing other data owners from reconstructing the transaction image based on all its features, thus effectively protecting the privacy of the transaction image. Simultaneously, based on some features of the transaction image, the digital watermark image embedded with the digital watermark is pre-simulated, thereby extracting the target evidence set from the digital watermark image to determine whether the image to be identified is a leaked image. This avoids the problem of falsely reporting leaks, achieving effective identification of leaked data and improving data security.
[0063] Embodiments of the present invention provide a data leakage detection method and terminal for intelligent connected vehicle data, which can be applied in blockchain data transaction scenarios. It can effectively identify leaked data while ensuring the privacy of transaction data, thereby improving the security of transaction data. Specific embodiments are described below:
[0064] Please refer to Figures 1 to 3 Embodiment 1 of the present invention is as follows:
[0065] A data leakage detection method for intelligent connected vehicle data is applied to the data leakage identification terminal for intelligent connected vehicle data. Specifically, such as... Figure 2As shown, the transaction scenario of this data breach identification terminal also includes a data owner, a data buyer, and a smart contract, all of which are connected to a blockchain network. The data breach identification terminal is the terminal corresponding to the data breach capture and evidence provider (e.g., users of a crowdsourcing platform). It identifies and obtains leaked data on an internet sharing platform, collaborates with the data owner to collect relevant evidence of the data breach, and finally submits the evidence to the smart contract for arbitration to obtain a reward. The data owner is the terminal corresponding to the owner of the intelligent connected vehicle data (e.g., car owners). It provides the intelligent connected vehicle data content and has the final decision-making power regarding the privacy, transaction, and sharing of intelligent connected vehicle data. The data purchaser is the terminal corresponding to the user of intelligent connected vehicle data (such as car manufacturers, car service providers, traffic management departments, research institutions, etc.). They acquire and use relevant content of intelligent connected vehicle data from the data owner, but must comply with laws and regulations related to data privacy protection. They cannot arbitrarily disclose raw data without the data holder's consent or after data anonymization. They also deposit a security deposit with the smart contract to facilitate accountability for data breaches. The smart contract is the terminal responsible for arbitration. It stores evidence provided by the other three parties (i.e., the data owner, the data purchaser, and the data breach identification party) and verifies the evidence. If the evidence is valid, the whistleblower is rewarded, and the party responsible for the breach is held accountable.
[0066] The data leakage detection method provided by this invention can detect whether relevant binary stream data is leaked. Specifically, when the data leakage identification terminal captures suspected leaked binary stream data on a public platform and wishes to earn a reward for reporting it, it collaborates with the data owner to collect leakage evidence and submits the evidence to a smart contract for arbitration. If the evidence is valid, the smart contract pays a reward to the data leakage identification terminal, and simultaneously deducts the deposit of the data purchaser to hold the data purchaser accountable for the leaked data. In this embodiment, the binary stream data is a digital image, that is, both the image to be identified and the original image are binary stream data.
[0067] like Figure 1 and Figure 3 As shown, the method includes:
[0068] S1. Obtain the image to be identified from a public platform and extract the target feature components of the image to be identified at different frequencies.
[0069] It should be noted that the images to be authenticated obtained by the data breach detection system on public platforms are unencrypted images. This is because images allowed to be publicly disclosed in a trading environment are those encrypted with the private key by the data purchaser. Other users, without the private key, cannot decrypt or identify the images, and therefore cannot use them for any other purpose to infringe on the data owner's rights or privacy. Thus, when a data purchaser leaks images to other users, it must decrypt them with its private key before disclosing them. Therefore, if the data breach detection system identifies an unencrypted image on a public platform, it can identify it as the image to be authenticated.
[0070] Specifically, in step S1, extracting the target feature components of the image to be identified at different frequencies includes:
[0071] S11. Perform discrete wavelet transform on the image to be identified to obtain the low-frequency feature components and high-frequency feature components of the image to be identified.
[0072] S12. Select low-frequency target feature components and high-frequency target feature components from the low-frequency feature components and the high-frequency feature components.
[0073] In some embodiments, the image to be identified, I'[i], is decomposed into low-frequency feature components L' using a three-level discrete wavelet transform (DWT). i and 3 high-frequency characteristic components H' i ,Right now Among them, the low-frequency characteristic component L' i =c1, high-frequency characteristic component H' i =(H' i 1,H' i 2,H' i 3). Specifically, the third-level high-frequency feature component is selected as the high-frequency target feature component of the image to be identified, thereby obtaining the low-frequency target feature component of the image to be identified as L'. i The high-frequency target feature component is H' i 3.
[0074] It should be noted that since the watermark feature components of images that are allowed to be publicly disclosed in the trading environment are hidden in the image's feature components, in order to recover the watermark feature components from the image to be identified, it is necessary to decompose the corresponding target feature components from the image to be identified.
[0075] S2. Send the image to be identified and the target feature component to the data owner of the image to be identified, in order to request the watermark embedding data corresponding to the target feature component in the original image corresponding to the image to be identified.
[0076] In some embodiments, the data leakage identification terminal sends relevant information about the image to be identified to the data owner terminal so that the data owner terminal can quickly determine the original image corresponding to the image to be identified.
[0077] In some embodiments, after receiving a request from the data leakage identification terminal, the data owner will first verify the authenticity of the request to prevent other user terminals from attacking the data owner through the request and to ensure the data security of the data owner.
[0078] Specifically, the watermark embedding data includes low-frequency original feature components in the original image corresponding to the low-frequency target feature components, high-frequency original feature components in the original image corresponding to the high-frequency target feature components, and embedding weight parameters in the original image corresponding to the low-frequency feature components and the high-frequency feature components when embedding digital watermarks.
[0079] In some embodiments, the low-frequency target feature component L' in the original image i The corresponding low-frequency original feature component is L i In the original image, the high-frequency target feature component H' i The corresponding high-frequency original feature component 3 is H. i 3. The original image, when embedding a digital watermark, is associated with the low-frequency feature component L'. i The corresponding embedding weight parameter is a1, and the original image is embedded with the high-frequency feature component H' when embedding the digital watermark. i =(H' i 1,H' i 2,H' i 3) The corresponding embedding weight parameters are a2, a3, and a4.
[0080] S3. Reconstruct the digital watermark image of the image to be identified based on the watermark embedding data, and extract the target evidence set from the digital watermark image.
[0081] Specifically, in step S3, restoring the digital watermark image of the image to be identified based on the watermark embedding data includes:
[0082] S31. Generate the low-frequency watermark feature component of the image to be identified based on the low-frequency target feature component, the low-frequency original feature component, and the embedding weight parameter corresponding to the low-frequency feature component.
[0083] In this embodiment, the data owner embeds the digital watermark image into the original image using a weighted method. Therefore, step S31 specifically involves: l' i =(L' i -L i ) / a1, where l' iL' represents the low-frequency watermark feature components of the image to be identified. i L represents the low-frequency target feature components of the image to be identified. i a1 represents the low-frequency original feature component of the original image, and a1 represents the low-frequency feature component L'. i The corresponding embedding weight parameters.
[0084] S32. Generate the high-frequency watermark feature component of the image to be identified based on the high-frequency target feature component, the high-frequency original feature component, and the embedding weight parameter corresponding to the high-frequency feature component.
[0085] In this embodiment, the data owner embeds the digital watermark image into the original image using a weighted method. Therefore, step S32 specifically involves: h' i =(H' i 3-H i ) / [a2,a3,a4], where h' i H' represents the high-frequency watermark feature components of the image to be identified. i 3 represents the high-frequency target feature components of the image to be identified, H' i 3=(cH3,cV3,cD3), H i H represents the high-frequency original feature components of the original image. i =(H i 1,H i 2,H i 3), a2, a3, and a4 represent the embedding weight parameters corresponding to the high-frequency target feature components cH3, cV3, and cD3, respectively.
[0086] It should be noted that, since the data owner does not need to disclose all the feature components of the entire image to the data leakage identification end during the process of extracting watermark feature components in the auxiliary data leakage identification end, the data leakage identification end cannot recover the original image based on the partial feature components of the entire image provided by the data owner, thus effectively protecting the privacy of the original transaction image.
[0087] S33. Reconstruct the digital watermark image of the image to be identified based on the low-frequency watermark feature components and the high-frequency watermark feature components.
[0088] Specifically, step S33 includes:
[0089] S331. Perform inverse wavelet transform on the low-frequency watermark feature components and the high-frequency watermark feature components to obtain the digital watermark image of the image to be identified.
[0090] In some embodiments, step S331 specifically includes: Where IDWT represents the inverse wavelet transform operation, and W' represents the digital watermark image.
[0091] Specifically, in step S3, extracting the target evidence set from the digital watermark image includes:
[0092] S34. Calculate the average pixel value of the digital watermark image, and generate a target evidence set based on the pixels in the digital watermark image whose pixel values are greater than the average pixel value.
[0093] In some embodiments, a target evidence set J' is determined based on the extraction rules of the standard evidence set J. Specifically, if the standard evidence set J stores standard pixels in the standard digital watermark image whose pixel values are less than or equal to the average pixel value, then pixels in the current digital watermark image W' whose pixel values are less than or equal to the average pixel value should be selected to generate the target evidence set J'. If the standard evidence set J stores standard pixels in the standard digital watermark image whose pixel values are greater than the average pixel value, then step S34 is executed.
[0094] S4. Verify the validity of the target evidence set through a smart contract. If the target evidence set is valid, then determine that the image to be identified is a leaked image.
[0095] Specifically, step S4 includes:
[0096] S41. Obtain the pre-stored standard evidence set of the original image through a smart contract, and calculate the intersection between the standard evidence set and the target evidence set;
[0097] S42. If the intersection is less than a preset standard value, then the target evidence set is determined to be invalid.
[0098] S43. If the intersection is not less than the standard value, then the target evidence set is determined to be valid.
[0099] Specifically, after determining that the target evidence set is valid, the method further includes:
[0100] S421. Mark the standard evidence set corresponding to the target evidence set.
[0101] Before calculating the intersection between the standard evidence set and the target evidence set, the method further includes:
[0102] S401. Detect whether the standard evidence set contains the marker. If it does, exit the verification and end the data leakage detection. If it does not, perform the step of calculating the intersection between the standard evidence set and the target evidence set.
[0103] Specifically, if the target evidence set is valid, the method further includes:
[0104] S5. In response to the payment instruction of the smart contract, receive the digital currency corresponding to the payment instruction and send the digital currency to the target account that uploaded the image to be identified.
[0105] The method also includes:
[0106] S6. If the target evidence set is invalid, a deduction instruction is sent to the target account to request the deduction of a preset amount of digital currency from the target account.
[0107] It should be noted that digital currency refers to legal tender similar to digital yuan issued by the central bank.
[0108] In some embodiments, steps S5 and S6 specifically involve: if the target evidence set provided by the user is valid, the smart contract will pay the user a digital currency (i.e., digital fiat currency, such as digital RMB) as a reward; if the target evidence set provided by the user is invalid, the smart contract will request the user to pay a digital currency to cover the cost of calling the smart contract.
[0109] In some embodiments, for a valid set of standard evidence, the smart contract will hold the data purchaser accountable. Specifically, this accountability involves deducting the digital currency pledged by the data purchaser as collateral. If the data owner wishes to further pursue the data purchaser's liability, they can utilize blockchain to trace the entire history of digital asset transactions, generating electronic evidence to pursue legal action against the data purchaser for privacy violations and other infringing issues.
[0110] Please refer to Figure 4 Embodiment two of the present invention is as follows:
[0111] The data leakage detection terminal 100 for intelligent connected vehicle data includes a memory 101, a processor 102, and a computer program stored on the memory 101 and running on the processor 102. When the processor 102 executes the computer program, it implements each step of the data leakage detection method for intelligent connected vehicle data described above.
[0112] In summary, the data leakage detection method and terminal for intelligent connected vehicle data provided by this invention, when the data leakage identification terminal identifies an unencrypted image to be identified on a public platform, extracts target feature components of the image to be identified at different frequencies, and then requests watermark embedding data from the data owner based on the target feature components. This watermark embedding data is then used to reconstruct the digital watermark image that may correspond to the image to be identified. Since the target feature components of this invention are only a portion of the feature components in the image to be identified, the data owner does not need to provide all feature components to the data leakage identification party, avoiding other data parties from reconstructing the transaction image based on the feature components presented at the time of evidence presentation, effectively ensuring the privacy of the transaction image. Simultaneously, after reconstructing the digital watermark image, the data leakage identification terminal extracts the target evidence set from the digital watermark image and verifies the similarity between the target evidence set and the standard evidence set through a smart contract, thereby determining whether the target evidence is valid and solving the problem of false reporting of leakage, achieving effective detection of leaked data. Furthermore, users who provide valid evidence are rewarded to incentivize more users to actively provide data leakage verification, improving data security; users who provide invalid evidence are penalized, reducing the cost of detecting leaked data.
[0113] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.
Claims
1. A data leakage detection method for intelligent connected vehicle data, characterized in that, include: The image to be identified is acquired from a public platform, and the target feature components of the image to be identified at different frequencies are extracted. The image to be identified and the target feature component are sent to the data owner of the image to be identified in order to request the watermark embedding data corresponding to the target feature component in the original image corresponding to the image to be identified. The digital watermark image of the image to be identified is reconstructed based on the watermark embedding data, and the target evidence set is extracted from the digital watermark image. The validity of the target evidence set is verified by a smart contract. If the target evidence set is valid, the image to be identified is determined to be a leaked image.
2. The data leakage detection method for intelligent connected vehicle data according to claim 1, characterized in that, The extraction of target feature components at different frequencies of the image to be identified includes: Perform discrete wavelet transform on the image to be identified to obtain the low-frequency feature components and high-frequency feature components of the image to be identified; Select low-frequency target feature components and high-frequency target feature components from the low-frequency feature components and the high-frequency feature components.
3. The data leakage detection method for intelligent connected vehicle data according to claim 2, characterized in that, The watermark embedding data includes low-frequency original feature components in the original image corresponding to the low-frequency target feature components, high-frequency original feature components in the original image corresponding to the high-frequency target feature components, and embedding weight parameters in the original image corresponding to the low-frequency feature components and the high-frequency feature components when embedding digital watermarks. The step of restoring the digital watermark image of the image to be identified based on the watermark embedding data includes: The low-frequency watermark feature components of the image to be identified are generated based on the low-frequency target feature components, the low-frequency original feature components, and the embedding weight parameters corresponding to the low-frequency feature components. The high-frequency watermark feature components of the image to be identified are generated based on the high-frequency target feature components, the high-frequency original feature components, and the embedding weight parameters corresponding to the high-frequency feature components. The digital watermark image of the image to be identified is reconstructed based on the low-frequency watermark feature components and the high-frequency watermark feature components.
4. The data leakage detection method for intelligent connected vehicle data according to claim 3, characterized in that, The step of reconstructing the digital watermark image of the image to be identified based on the low-frequency watermark feature components and the high-frequency watermark feature components includes: The low-frequency watermark feature components and the high-frequency watermark feature components are subjected to inverse wavelet transform to obtain the digital watermark image of the image to be identified.
5. The data leakage detection method for intelligent connected vehicle data according to claim 1, characterized in that, The extraction of the target evidence set from the digital watermarked image includes: Calculate the average pixel value of the digital watermark image, and generate a target evidence set based on the pixels in the digital watermark image whose pixel values are greater than the average pixel value.
6. The data leakage detection method for intelligent connected vehicle data according to claim 1, characterized in that, The verification of the validity of the target evidence set via smart contract includes: The standard evidence set pre-stored in the original image is obtained through a smart contract, and the intersection between the standard evidence set and the target evidence set is calculated. If the intersection is less than a preset standard value, the target evidence set is determined to be invalid. If the intersection is not less than the standard value, then the target evidence set is determined to be valid.
7. The data leakage detection method for intelligent connected vehicle data according to claim 6, characterized in that, After determining that the target evidence set is valid, the process also includes: Mark the standard evidence set corresponding to the target evidence set; Before calculating the intersection between the standard evidence set and the target evidence set, the method further includes: Detect whether the marker exists in the standard evidence set; If it exists, exit the verification and end the data breach detection; If it does not exist, then perform the step of calculating the intersection between the standard evidence set and the target evidence set.
8. The data leakage detection method for intelligent connected vehicle data according to claim 1, characterized in that, If the target evidence set is valid, it also includes: In response to the payment instruction of the smart contract, the system receives the digital currency corresponding to the payment instruction and sends the digital currency to the target account that uploaded the image to be authenticated.
9. The data leakage detection method for intelligent connected vehicle data according to claim 8, characterized in that, Also includes: If the target evidence set is invalid, a deduction instruction is sent to the target account to request the deduction of a preset amount of digital currency from the target account.
10. A data leakage detection terminal for intelligent connected vehicle data, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements each step of the data leakage detection method for intelligent connected vehicle data as described in any one of claims 1-9.