Electronic Document Authenticity Verification System
The system addresses hash instability and personal information exposure by converting documents to static images with dual watermarks and blockchain registration, ensuring secure and reliable authenticity verification.
Patent Information
- Application Number
- KR1020250057882
- Authority / Receiving Office
- KR · KR
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2025-05-01
- Publication Date
- 2026-11-10
AI Technical Summary
Existing methods for verifying electronic document authenticity face issues with hash value instability, sensitivity to user changes, dynamic document structures, exposure of personal information, and unintuitiveness, leading to unreliable and insecure document verification.
An electronic document authenticity verification system that converts documents to fixed static images, embeds dual watermarks and metadata, and registers hash values on a blockchain ledger, ensuring document integrity and privacy through decentralized verification.
Ensures consistent hash values, robust forgery detection, protects personal information, and provides intuitive verification, while maintaining document security and reliability.
Smart Images

Figure PAT00019_ABST
Abstract
Description
Technology Field
[0001] The present invention relates to the verification of the authenticity of electronic documents, and in particular, to an electronic document authenticity verification system capable of simultaneously achieving the protection of personal information and the prevention of forgery and alteration of electronic documents. Background Technology
[0002] In modern society, electronic documents have established themselves as an important means of replacing paper documents and are being utilized in various fields, such as contracts, certificates, proofs, and transaction statements.
[0003] Such electronic documents require a high level of security and authenticity throughout the entire process of creation, transmission, storage, and verification, and existing methods for verifying electronic documents are based on the following methods.
[0004] 1. Electronic signature or digital certificate-based method
[0005] It is a method of signing and verifying electronic documents using a Public Key Infrastructure (PKI).
[0006] However, it is difficult for users to directly determine whether the signer information or the document has been forged or altered, and verification itself is impossible if the signature certificate has expired or is invalid.
[0007] 2. Hash value-based blockchain registration method
[0008] It is a method that verifies whether a document has been altered by generating a hash value based on the content of the document and registering it on the blockchain.
[0009] However, most document formats, such as PDF and DOCX, have dynamic internal structures, and there is a problem where internal metadata (modification date, viewer information, etc.) changes and the hash value changes simply by viewing or saving.
[0010] As a result, the document may be perceived as forged regardless of the user's intent, leading to reduced reliability.
[0011] 3. Server-based log storage and comparison method
[0012] It is a method of verifying the authenticity of a document by comparing it with logs or history stored on a server.
[0013] This method involves a structure that stores document content on a central server or leaves comparison data, posing a high risk of personal information exposure and raising concerns about user privacy infringement.
[0014] The main problems with the existing technology described above are as follows.
[0015] (1) Instability of hash values
[0016] Even minor changes in the internal structure affect the hash value, so it is common for documents with identical content to have different hash values.
[0017] (2) Dynamic characteristics of the document
[0018] For formats such as PDF or DOCX, there are cases where the hash value changes even though the content has not changed, simply due to the user's viewing or changes to print settings.
[0019] (3) Sensitivity to unintended user changes
[0020] There are cases where the determination of document authenticity operates excessively sensitively, resulting in even legitimate user actions being judged as forgery.
[0021] (4) Concerns about exposure of personal information and original files
[0022] If original documents are stored on a server or logs are recorded to verify authenticity, there is a possibility of security breaches and personal information leakage.
[0023] (5) Unintuitiveness of the verification procedure
[0024] General users face the inconvenience of not being able to easily verify the authenticity of documents and having to understand a complex certificate system.
[0025] Therefore, the existing methods described above have limitations in terms of preventing document forgery or alteration and protecting personal information, and there is a problem in that they are difficult to function as a "fixed electronic document authenticity verification system" in the true sense. Prior art literature
[0026] (Patent Document 0001) KR 10-2016-0150278 A(Patent Document 0002) KR 10-1954268 B1(Patent Document 0003) KR 10-2018-0110670 A(Patent Document 0004) KR 10-2241246 B1(Patent Document 0005) KR 10-2733150 B1 The problem to be solved
[0027] The objective of the present invention, which is to solve the problems derived from the prior background technology, is to achieve reliability, security, privacy protection, and ease of use of electronic documents through four core technical elements: 'fixed image-based electronic document representation', 'double watermark insertion', 'blockchain-based decentralized authenticity verification', and 'server-independent structure'.
[0028] Meanwhile, the objectives of the present invention are not limited to those mentioned above, and other unmentioned objectives will be clearly understood from the description below. means of solving the problem
[0029] The above objective can be achieved by an electronic document authenticity verification system characterized by comprising: an electronic document generation module that, according to an embodiment of the present invention, receives electronic document data and renders it into a visual document format, converts the rendered document into a static image file in a fixed form, embeds watermark information and metadata to be used for identification and authenticity verification of the electronic document within the static image file, calculates the hash value of the final static image file in which the watermark information and metadata are embedded, and records and registers the hash value and related verification information in a blockchain ledger; and an electronic document authenticity verification module that receives a static image file for which authenticity verification is requested, extracts watermark information and metadata embedded within the file, queries verification information registered in the blockchain ledger in the electronic document generation module using the extracted information or the current hash value of the received static image file, and compares the extracted information and the current hash value of the received static image file with the queried blockchain verification information to determine the authenticity of the received static image file.
[0030] Preferably, the electronic document generation module comprises: a watermark processing unit that generates an image file in a format capable of lossless compression as the static image file, generates a high-frequency watermark value and a low-frequency watermark value derived from the original data of the electronic document, and inserts the high-frequency watermark value into the least significant bit (LSB) region of the pixel data of the image file and the low-frequency watermark value by changing the coefficients of the frequency region obtained by the Discrete Cosine Transform (DCT); and a metadata processing unit that hashes metadata representing the uniqueness and publication information of the electronic document using a unidirectional hash function and embeds the hashed metadata inside the image file.
[0031] Preferably, the electronic document authenticity verification module comprises: a data extraction unit that extracts embedded watermark information and hashed metadata from the input image file; an information comparison unit that, using the information extracted by the data extraction unit or the current hash value of the input image file, queries cache information for lookup (information stored by the electronic document generation module after blockchain registration) and verification information on the blockchain ledger cached in a separate database (DB) and compares the queried cache information and blockchain verification information with the extracted information and the current hash value; and an authenticity determination unit that determines the input static image file as authentic only when the extracted watermark information, the extracted hashed metadata, and the current hash value all match the verification information on the blockchain ledger, based on the comparison result of the information comparison unit.
[0032] Preferably, the electronic document generation module is the following mathematical formula
[0033]
[0034] Generating an electronic document by (where M represents the final data of the electronic document to be registered on the blockchain, and H SHA256 is the SHA256 solution function, θ mata is a metadata insertion function, θ watermark is a watermark insertion function, μ is the metadata set of the electronic document, ω DCT is the low-frequency watermark value, ω LSB It is characterized by a high-frequency watermark value, ω is data for creating the watermark, and I is the byte of the rendered electronic document, i.e., the image file.
[0035] Preferably, the electronic document authenticity verification module is the following mathematical formula
[0036]
[0037] Verifying an electronic document by (where D refers to an electronic document uploaded by a user, μ′ refers to metadata extracted from electronic document D, and μ db is metadata cached in the DB, and μ bc refers to metadata registered on the blockchain, and represents the watermark hash value extracted from electronic document D, and refers to the watermark hash value registered on the blockchain, and is the SHA256 hash value of the entire electronic document D and It is characterized by the hash value of electronic document D registered on the blockchain. Effects of the invention
[0038] According to the present invention based on the above-described embodiment, the authenticity of an electronic document can be verified and forgery and alteration prevented, while simultaneously realizing privacy protection, and accordingly, the following specific technical effects are provided.
[0039] 1. Ensuring the security of hash values
[0040] The present invention ensures consistency between the visual representation of a document and its hash value by converting an electronic document into an image format with a fixed visual structure (e.g., a PNG image) rather than existing dynamic formats such as PDF or DOCX.
[0041] Accordingly, the same hash value is always generated for identical document content, and hash inconsistency issues caused by simple viewing or saving do not occur.
[0042] 2. Enhanced forgery detection capabilities through dual watermark insertion
[0043] By inserting watermarks into the high-frequency range (LBS method) and the low-frequency range (DCT conversion method), respectively, the authenticity of the document can be detected directly from the document itself without complex image manipulation.
[0044] Since watermarks embedded in documents are based on unique information such as UID, Publisher ID, and timestamps, tampering or forgery breaks watermark consistency and can be detected during the verification process.
[0045] 3. Authenticity Verification and Prevention of Tampering through Blockchain Integration
[0046] By registering hash values generated from images such as PNG format on the blockchain, the authenticity of documents can be verified on a public distributed ledger without relying on a central server.
[0047] Blockchain-based registration confirms the authenticity of the document at the time of creation and allows for easy detection of any subsequent changes.
[0048] 4. Realization of personal information protection through a structure that stores original files and does not retain logs
[0049] The present invention adopts a structure that does not store original documents or usage logs on the server, so that the user's sensitive information is not exposed externally.
[0050] It temporarily caches only the minimum information required for verification, and this information consists of non-identifiable metadata to maximize privacy protection.
[0051] 5. Improvement of the intuitiveness and ease of use of the verification process
[0052] Authenticity can be verified using only a single image file, without the need for a separate digital signature verification process or certificate.
[0053] Users can quickly determine whether a document has been forged or tampered with through a verification tool that automatically analyzes watermarks and metadata embedded in the document, even without specialized knowledge.
[0054] 6. Ease of application in various practical environments
[0055] The present invention is applicable to various types of electronic documents, such as financial transaction documents, delivery slips, electronic contracts, certificates of employment and degree, and electronic tax invoices, and has high practicality that can be used immediately in practice.
[0056] Watermarks and metadata embedded in documents can be used as evidence in disputes, and any forgery or alteration can be immediately identified due to discrepancies in the watermarks.
[0057] 7. Simultaneous assurance of security and reliability
[0058] Through a triple structure of visual fixation + watermark hiding + blockchain hash fixation, the appearance of the document is maintained while its authenticity and whether it has been tampered with can be verified using advanced technology, thereby ensuring both security and reliability.
[0059] In conclusion, the present invention has high technical effectiveness in that it technically resolves problems associated with existing electronic document verification technologies, such as structural instability, hash value inconsistencies, limitations of centralized storage, and infringement of user privacy, and provides an electronic document verification system that satisfies prevention of forgery and tampering, authentication of authenticity, and protection of personal information. Brief explanation of the drawing
[0060] FIG. 1 is a block diagram showing an electronic document authenticity verification system according to an embodiment of the present invention, and FIG. 2 is a flowchart of an electronic document generation process according to an embodiment of the present invention, and FIG. 3 is a flowchart of an electronic document authenticity verification process according to an embodiment of the present invention. Specific details for implementing the invention
[0061] The advantages and features of the present invention, and the methods for achieving them, will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but may be implemented in various different forms; these embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention. Furthermore, the terms used in this specification are for describing the embodiments and are not intended to limit the present invention. In this specification, the singular form includes the plural form unless specifically stated otherwise in the text.
[0062] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings. Meanwhile, the illustration and detailed description of configurations and their operations and effects that are easily known to those skilled in the art will be simplified or omitted, and the details will be described focusing on the parts related to the present invention.
[0063] As illustrated in FIG. 1, the present invention is largely composed of an electronic document generation module and an electronic document authenticity verification module. It converts electronic documents into fixed static image files, inserts high-frequency (LSB) and low-frequency (DCT) watermarks and metadata, and registers relevant information in a blockchain ledger to guarantee authenticity. Additionally, it determines authenticity by extracting watermarks and metadata from documents submitted by users and comparing them with database and blockchain information. The present invention provides a comprehensive technical framework applicable to various documents and industries, without being limited to specific document types, technical implementation methods, or application environments. Below, the components, operation procedures, data processing methods, technical features, and applicability of each module are described in detail, and the entire flow from the generation to the verification of electronic documents is systematically explained with reference to FIG. 2 and FIG. 3.
[0064] First, referring to FIG. 2, the electronic document generation module receives electronic document data, renders it into a visual document format, converts the rendered document into a static image file in a fixed form, embeds watermark information and metadata to be used for identification and verification of the authenticity of the electronic document within the static image file, calculates the hash value of the final static image file with the watermark information and metadata embedded, and performs the function of recording and registering the hash value and related verification information in the blockchain ledger.
[0065] Here, it is preferable that the electronic document generation module generates an image file in a format capable of lossless compression as the static image file.
[0066] This electronic document generation module includes a watermark processing unit that generates high-frequency and low-frequency watermark values derived from the original data of the electronic document. In this case, the high-frequency watermark value is inserted into the Least Significant Bit (LSB) region within the pixel data of the image file, and the low-frequency watermark value is inserted by modifying the coefficients in the frequency domain of the Discrete Cosine Transform (DCT).
[0067] And the electronic document generation module includes a metadata processing unit that hashes metadata representing the uniqueness and issuance information of the electronic document using a one-way hash function and embeds the hashed metadata inside the image file.
[0068] The steps performed in the above electronic document generation module are comprehensively explained as follows.
[0069] 1. Enter document information
[0070] The creation of electronic documents begins with the process of receiving core document data from the user. The input data includes various items depending on the purpose and content of the document, encompassing unique identifiers (UIDs), issuer information, issuance date, document type, recipient information, transaction amount, contract terms, item lists, and certification details. For example, a contract includes the contracting parties, date of signing, amount, and conditions; an invoice includes the issuer's company name, recipient address, items, and amount. A certificate includes the issuing institution, certificate type, and recipient information, while a transaction statement collects transaction details, dates, and amounts. This data is entered in a structured format and supports text, numbers, dates, or other data formats as needed.
[0071] The input process is flexibly designed to accommodate diverse user environments. Users can directly input data for a single document or utilize standardized file formats to process large volumes of documents in batches. Validation is performed to ensure the accuracy of input data, and correction requests are sent to the user if missing required fields, format errors, or data exceeding ranges are detected. For example, UIDs are restricted to non-duplicate strings to guarantee uniqueness, and issue dates must adhere to international standard time formats. Furthermore, multilingual support allows global users to input data in various languages, and an intuitive user interface is provided to enhance accessibility.
[0072] 2. Visual Document Rendering
[0073] Based on input data, the system generates visual document formats by inserting the data into predefined templates. Templates are designed in various ways according to the purpose of the document and industry-specific requirements, and encompass all types of electronic documents, such as contracts, invoices, certificates, transaction statements, official documents, qualification certificates, employment certificates, and medical records. The elements defining the visual structure and content of the document in the template may consist of the following examples.
[0074] - Layout: Based on the document size (e.g., A4, A3, Letter), issuer information and the logo may be placed at the top, key content (e.g., contract terms, item list, certification information) in the center, and the issue date, document number, and certification phrase at the bottom. Margins, line spacing, and alignment may be adjusted according to the document type.
[0075] - Visual Elements: Standard typefaces, font sizes, and color palettes are used to ensure consistent output. For example, headings are emphasized with larger fonts, and body text is set to an appropriate size to enhance readability. Colors are primarily black and white or solid, but logos, certification marks, and highlighted text may be displayed in specific colors.
[0076] - Additional elements: May include the publisher logo, certification mark, barcode, QR code, unique document number, page number, etc. The QR code includes a verification URL to enable users to quickly verify authenticity.
[0077] The rendering process is performed through precise mapping between input data and templates. For example, the item list on an invoice is arranged in a table format, and if there are many items, it is split into multiple pages. Contracts clearly separate and arrange the terms and signature fields, while certificates highlight the issuing authority's logo and certification text. Rendering includes automatic adjustment features to prevent text overflow, image distortion, or layout errors, and is designed to always generate the same visual output for the same input data. This ensures the visual consistency of the document and the stability of subsequent hash value calculations.
[0078] 3. Create static image file
[0079] The rendered visual document is converted into a fixed static image file. The present invention uses a format capable of lossless compression, which fixes the visual representation of the document to prevent changes in content or metadata during viewing, storage, and transmission. During the conversion process, the resolution of the document is appropriately set according to the document size and purpose, and the color supports black and white, monochrome, or multicolor modes. For example, simple documents such as certificates can be converted to black and white mode to minimize file size, while contracts or invoices can use multicolor mode to include logos and emphasis elements.
[0080] Static image files balance quality and size by adjusting compression levels, and unnecessary metadata is removed. This ensures that identical image data is always generated for the same document content, maintaining consistency in subsequent hash value calculations. For multi-page documents, each page can be generated as a separate image file or processed as a single integrated file, with linkage information between pages included in the metadata. Image files undergo integrity checks to prevent corruption, and a regeneration procedure is initiated if an error is detected.
[0081] Here, static image files are preferably in the PNG (Portable Network Graphics) format, which is a lossless compression algorithm that ensures no data or quality is lost when compressing images.
[0082] 4. Insert Watermark Information
[0083] The watermark processing unit inserts high-frequency and low-frequency watermarks into image files. The watermarks are generated based on the document's Unique User ID (UID), issuer information, issue date, document type, and other related data, and are used for determining authenticity and detecting forgery or tampering.
[0084] - High-Frequency Watermark: Watermark data is inserted by introducing subtle changes to the image's pixel data. This method minimizes visual distortion, preventing users from perceiving the changes with the naked eye, and the watermark data is accurately extracted during the verification process. High-frequency watermarks can be inserted across specific areas or the entire document, and the data volume and insertion location are adjusted based on the document's size and content.
[0085] - Low-frequency watermark: By embedding data into the frequency domain of an image, it provides robustness to ensure the watermark persists even under various transformations such as resizing, compression, scaling, and rotation. This method can be concentrated in the main content area of a document or distributed across the entire image.
[0086] The above watermark data represents the uniqueness and publication information of the document and is protected from unauthorized modification or extraction from the outside. The watermark can be designed to include an error correction mechanism to enable data restoration even from damaged images.
[0087] The present invention achieves both concealment and toughness simultaneously by combining high-frequency and low-frequency watermarks, and provides flexibility suitable for various document types and transformation environments.
[0088] 5. Insert metadata
[0089] The metadata processing unit generates metadata representing the uniqueness and issuance information of an electronic document and inserts it into an image file. This metadata may include various information such as UID, issuer information, issuance date, document type, document version, issuance location, and document category. This data is hashed using a one-way hash function and then recorded in the metadata storage area of the image file. The hashed metadata cannot be decrypted and is compared with information in the blockchain ledger during the verification process.
[0090] Metadata insertion ensures document integrity and traceability, and reduces dependency on external databases by including all validation information within the image file itself. For multi-page documents, page-specific metadata or integrated metadata representing the entire document may be inserted. The size of the metadata is adjusted according to the document's purpose and requirements, and compression techniques are applied when necessary to enhance efficiency.
[0091] 6. Calculate hash value
[0092] Image files embedded with watermarks and metadata have their hash values calculated using a one-way hash function. This hash value serves as a unique identifier for the document, and a different hash value is generated if the file's content is altered. The calculation process applies to the entire data of the image file and is performed with the watermark and metadata included. For multi-page documents, partial tampering can be detected by generating page-specific hash values and a combined hash value for the entire document. The hash value guarantees the integrity of the document and is subsequently registered in the blockchain ledger.
[0093] 7. Registration of Blockchain Ledger
[0094] The calculated image file hash value, hashed metadata, watermark data hash value, and related information are registered in the blockchain ledger. The blockchain guarantees the authenticity of documents by leveraging the immutability of distributed ledgers, making it impossible to tamper with or forge the registered data. The registered data includes the image file hash value, metadata hash value, watermark hash value, timestamp, and document type, and this data is combined and recorded as a single transaction. The transaction is processed using unidirectional hashing, making it impossible to restore the original information, and the transaction identifier returned after registration is referenced during the verification process.
[0095] The present invention supports various blockchain environments, and a public blockchain, private blockchain, or hybrid blockchain can be selected according to the security requirements of the documents. For example, public documents may be registered on a public blockchain for transparency, while sensitive contracts may be registered on a private blockchain for security. In the case of multiple documents, blockchain registration can be performed on an individual document or as a group of documents.
[0096] 8. Minimal Information Caching
[0097] After registration on the blockchain, the system caches minimal information in the database to enhance verification performance. Cached data is limited to transaction identifiers, hashed metadata, watermark hash values, and timestamps, and does not include the content of the original document or user identification information. This data is stored in a de-identified form so that the original data cannot be inferred upon external access. Cached information is automatically deleted after a certain period, and the database is used solely as an auxiliary reference for the blockchain ledger.
[0098] 9. Mathematical expression
[0099] The process of creating an electronic document is expressed by the following mathematical formula 1.
[0100]
[0101] Here, M represents the final data of the electronic document to be registered on the blockchain, and H SHA256 is the SHA256 solution function, θ mata is a metadata insertion function, θ watermark is a watermark insertion function, μ is the metadata set of the electronic document, ω DCT is the low-frequency watermark value, ω LSB is the high-frequency watermark value, ω is the data for creating the watermark, and I is the byte of the rendered electronic document, i.e., the image file.
[0102] Mathematical formula 1 as described above represents the process in which document data is converted into a hash value and registered on the blockchain after inserting a watermark and metadata.
[0103] 10. Distribution and Privacy Protection
[0104] The generated image files are securely distributed to users. Distribution is carried out via a time-limited, signed access path that expires after a certain period to block unauthorized access. The files are deleted from the system after distribution, and the original documents, user input data, or related logs are not stored. All data processing is performed in a de-identified form to protect user privacy and comply with relevant laws.
[0105] Next, referring to FIG. 3, the electronic document authenticity verification module receives a static image file for which authenticity verification is requested, extracts watermark information and metadata embedded within the file, queries verification information registered in the blockchain ledger of the electronic document generation module using the extracted information or the current hash value of the received static image file, and performs the function of determining the authenticity of the received static image file by comparing the extracted information and the current hash value of the received static image file with the queried blockchain verification information.
[0106] Here, the electronic document authenticity verification module includes a data extraction unit that extracts embedded watermark information and hashed metadata from the input image file.
[0107] And the electronic document authenticity verification module includes an information comparison unit that uses the information extracted by the data extraction unit or the current hash value of the input image file to query cache information for retrieval (information stored by the electronic document generation module after blockchain registration) and verification information on the blockchain ledger cached in a separate database (DB), and compares the queried cache information and blockchain verification information with the extracted information and the current hash value.
[0108] Additionally, the electronic document authenticity verification module includes an authenticity determination unit that determines the input static image file as authentic only when the extracted watermark information, the extracted hashed metadata, and the current hash value all match the verification information on the blockchain ledger, based on the comparison result of the information comparison unit.
[0109] The steps performed in the above electronic document authenticity verification module are comprehensively explained as follows.
[0110] 1. Submit image file
[0111] The verification process begins with the user submitting an image file to the system for authenticity verification. The submitted file is processed immediately by the system, and its format and integrity are verified. If the file is corrupted or not in the proper format, an error message is provided to the user, who is instructed to resubmit the correct file. The present invention supports various submission paths, and the user can submit a single file or multiple files.
[0112] 2. Extraction of Watermarks and Metadata
[0113] The data extraction unit extracts high-frequency watermarks, low-frequency watermarks, and metadata from the submitted image file.
[0114] - High-frequency watermark extraction: Recovers watermark data from image pixel data. This data contains the document's unique identification information (UID, publisher information, etc.) and is used for verification.
[0115] - Low-frequency watermark extraction: Extracts watermark data in the frequency domain of the image. Since this data is preserved even in modified images, authenticity can be verified even if the file has been modified or re-scaled.
[0116] - Metadata Extraction: Reads hashed metadata from the image file's metadata storage area. This metadata consists of hash values such as UID, publisher information, and publication date.
[0117] The extracted data is used in a subsequent comparison process, and if the data is missing or damaged, it is determined to be forged.
[0118] 3. Check for data existence
[0119] The presence of extracted watermarks and metadata is verified. If watermarks or metadata are missing or damaged, the document is considered forged, and the verification process is terminated. This step provides an initial assessment of the document's integrity to prevent unnecessary subsequent processing.
[0120] 4. Database Information Comparison
[0121] The information comparison unit compares the extracted watermark and metadata with information cached in the database. The cached information includes transaction identifiers, hashed metadata, and watermark hash values stored after blockchain registration upon the creation of the electronic document. The comparison process verifies whether the extracted data matches the cached data; if a discrepancy occurs, it is determined to be forgery. This step enhances efficiency through rapid querying.
[0122] 5. Comparison of Blockchain Ledger Information
[0123] Documents that pass the database comparison are finally compared with the information in the blockchain ledger. The information comparison unit queries the data registered in the blockchain using cached transaction identifiers. The retrieved data includes the hash values of image files, metadata, and watermarks. The comparison compares the extracted data with the blockchain data, and if a discrepancy occurs, it is determined to be forgery. Blockchain comparison guarantees reliability by leveraging the immutability of the distributed ledger.
[0124] 6. Determination of Authenticity
[0125] The Authenticity Verification Unit determines the authenticity of a document by synthesizing all comparison results. Documents that pass both database and blockchain comparisons are judged to be authentic, while those that do not are judged to be forged or unverifiable. The judgment results are presented to the user intuitively and include the document's UID, issue date, and a summary of the comparison results.
[0126] 7. Mathematical expression
[0127] The electronic document verification process is expressed by the following mathematical formula 2.
[0128]
[0129] Here, D represents an electronic document uploaded by a user, μ′ represents metadata extracted from electronic document D, and μ db is metadata cached in the DB, and μ bc refers to metadata registered on the blockchain, and represents the watermark hash value extracted from electronic document D, and refers to the watermark hash value registered on the blockchain, and is the SHA256 hash value of the entire electronic document D and represents the hash value of electronic document D registered on the blockchain.
[0130] Mathematical formula 2 as described above indicates that the data of a submitted document must match both the database and blockchain information to be determined to be authentic.
[0131] Meanwhile, during the aforementioned verification process, the system of the present invention does not store original documents or user-related information. Submitted image files are deleted immediately after verification is completed, and extracted data is also discarded. Information cached in the database is limited to a de-identified form, and the user's identity or document content cannot be inferred.
[0132] According to the invention described so far, the following specific effects can be realized.
[0133] 1. Improved hash stability by ensuring visual fixedness
[0134] The present invention allows the structure and visual representation of a document to be maintained consistently by rendering an electronic document as a fixed image.
[0135] Accordingly, the hash value does not change even if the user views the document or opens it with other software, and the same hash value can always be generated for documents with the same content.
[0136] This can fundamentally resolve the "unintentional hash value change" problem that occurred in existing PDF, DOCX, etc.
[0137] 2. Enhanced prevention of forgery and tampering and verification of authenticity through the insertion of a dual watermark (high frequency + low frequency).
[0138] The present invention enables highly reliable authenticity verification without visual damage by inserting a high-frequency watermark utilizing the lowermost bit (LSB) of a pixel together with a low-frequency (DCT-based) watermark that manipulates the frequency domain.
[0139] Furthermore, by combining multiple watermarking techniques, the possibility of forgery resulting from simple copying or image manipulation can be significantly reduced, and since authenticity can be detected directly from the image itself, independent verification is possible.
[0140] 3. Securing Decentralization and Impossibility of Forgery through Blockchain-based Verification
[0141] By registering the hash value and metadata of the generated PNG image on the blockchain, authenticity can be verified without a central authority.
[0142] Since a blockchain is a distributed ledger that cannot be forged or deleted, anyone can detect tampering if a document is altered after the hash value is registered. This structure can provide an irreversible verification mechanism for documents.
[0143] 4. Protection of personal information through non-retention of original documents and a non-logging structure
[0144] The present invention does not store original electronic documents or usage logs on the server, and utilizes only a minimal amount of anonymized information by temporarily caching it during verification. This structure blocks the possibility of monitoring or tracking user behavior and prevents sensitive document content or identifying information from being exposed externally, thereby enabling the realization of a structure that complies with the Personal Information Protection Act and the requirements for enhanced privacy.
[0145] 5. User-friendly authenticity verification process
[0146] Authenticity can be verified using only a simple image file (PNG) without the need for complex certificate installation or signature file verification.
[0147] General users can automatically extract watermarks and metadata through web-based verification services without a separate program, and verify authenticity by comparing them with the blockchain.
[0148] It can provide high convenience, allowing even users unfamiliar with technology to easily access and use it.
[0149] 6. Diverse Application Potential
[0150] The structure of the present invention is applicable to various forms of electronic documents such as invoices, electronic contracts, certificates, and proof documents, and can be utilized in various fields such as finance, law, public institutions, and education.
[0151] In particular, since hidden information such as timestamps, issuer information, and document unique numbers is included in the form of a watermark, it has the advantage of being usable as strong evidence in legal disputes related to document forgery.
[0152] The foregoing description has broadly detailed the features and technical advantages of the present invention to better facilitate a better understanding of the claims of the invention to be described below. Those skilled in the art will understand that the present invention may be implemented in other specific forms without altering its technical concept or essential features. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. The scope of the present invention is defined by the claims set forth below rather than by the detailed description above, and all modifications or variations derived from the claims and their equivalents should be interpreted as being included within the scope of the present invention.
Claims
Claim 1 An electronic document authenticity verification system characterized by comprising: an electronic document generation module that receives electronic document data, renders it into a visual document format, converts the rendered document into a static image file in a fixed form, embeds watermark information and metadata to be used for identification and authenticity verification of the electronic document within the static image file, calculates the hash value of the final static image file with the watermark information and metadata embedded therein, and records and registers the hash value and related verification information in a blockchain ledger; and an electronic document authenticity verification module that receives a static image file for which authenticity verification is requested, extracts watermark information and metadata embedded within the file, queries verification information registered in the blockchain ledger in the electronic document generation module using the extracted information or the current hash value of the received static image file, and compares the extracted information and the current hash value of the received static image file with the queried blockchain verification information to determine the authenticity of the received static image file. Claim 2 An electronic document authenticity verification system according to claim 1, wherein the electronic document generation module comprises: a watermark processing unit that generates an image file in a format capable of lossless compression as the static image file, generates a high-frequency watermark value and a low-frequency watermark value derived from the original data of the electronic document, and inserts the high-frequency watermark value into the least significant bit (LSB) region within the pixel data of the image file and the low-frequency watermark value by changing the coefficients in the frequency region of the discrete cosine transform (DCT); and a metadata processing unit that hashes metadata representing the uniqueness and publication information of the electronic document using a unidirectional hash function and embeds the hashed metadata inside the image file. Claim 3 An electronic document authenticity verification system according to claim 1 or 2, wherein the electronic document authenticity verification module comprises: a data extraction unit that extracts embedded watermark information and hashed metadata from the input image file; an information comparison unit that, using the information extracted by the data extraction unit or the current hash value of the input image file, queries cache information for querying cached in a separate database (DB) (information stored by the electronic document generation module after blockchain registration) and verification information on a blockchain ledger, and compares the queried cache information and blockchain verification information with the extracted information and the current hash value; and an authenticity determination unit that determines the input static image file as authentic only when the extracted watermark information, the extracted hashed metadata, and the current hash value all match the verification information on the blockchain ledger, based on the comparison result of the information comparison unit. Claim 4 In paragraph 1 or 2, the electronic document generation module is the following mathematical formula Generating an electronic document by (where M represents the final data of the electronic document to be registered on the blockchain, and H SHA256 is the SHA256 solution function, θ mata is a metadata insertion function, θ watermark is a watermark insertion function, μ is the metadata set of the electronic document, ω DCT is the low-frequency watermark value, ω LSB An electronic document authenticity verification system characterized by: is a high-frequency watermark value, ω is data for creating the watermark, and I is the byte of the rendered electronic document, i.e., the image file. Claim 5 In paragraph 1 or 3, the electronic document authenticity verification module is the following mathematical formula Verifying an electronic document by (where D refers to an electronic document uploaded by a user, μ′ refers to metadata extracted from electronic document D, and μ db is metadata cached in the DB, and μ bc refers to metadata registered on the blockchain, and represents the watermark hash value extracted from electronic document D, and refers to the watermark hash value registered on the blockchain, and is the SHA256 hash value of the entire electronic document D and An electronic document authenticity verification system characterized by (meaning the hash value of electronic document D registered on the blockchain).
