An AI proportion self-consistent deduction-based perfect mature state automatic restoration method and system for cultural relics

Abstract

The application discloses a kind of based on AI proportion self-consistent deduction relic perfect mature state automatic recovery method and system, it is related to relic digital protection and artificial intelligence technical field.The present application takes the perfect mature state of the stable structure of relic, proportion self-consistent, function complete as benchmark, through AI extraction residual component in the structure invariant that can be uniquely determined, eliminate deformation, corrosion etc.after disturbance, realize local to the structure continuity deduction of whole.For the face of relic Set quantitative recovery threshold, when residual face is not less than 1 / 4 and contains central axis or single side contour, can be recovered according to symmetry and combined with the shape of literature of the same period and the proportion of similar statue auxiliary recovery;It is not recovered below 1 / 4 or no effective anchor point, and the feature that cannot be uniquely determined remains the state of defect.The present application realizes objective, scientific, automatic relic structure recovery, and is suitable for the digital protection and repair of various material damaged relics.

Description

Technical Field

[0001] This invention relates to the fields of digital preservation of cultural relics, application of artificial intelligence algorithms, and restoration of cultural relics, specifically to a method based on... A method and system for the automatic restoration of cultural relics in a perfect, mature state based on AI-based proportional self-consistent extrapolation. Background Technology

[0002] Cultural relics commonly suffer from incompleteness, deformation, weathering, and damage during long-term burial and transmission. The restoration of cultural relics is a crucial aspect of cultural heritage preservation. A crucial aspect of artifact preservation. While the industry has long relied on experience based on proportion, symmetry, and ergonomics for manual restoration, all methods still adhere to... Restorers rely on subjective judgment, lacking standardized quantitative criteria, which can easily lead to fabricated restorations. Current digital restoration methods largely remain at the level of piecing together fragmented data. Compared with image completion, it lacks a holistic deductive system based on structural invariants and sets clear restoration thresholds, thus failing to achieve objective and accurate results. Repeatable and automated restoration of cultural relic structures. Therefore, a scientific, quantifiable, and automated method for restoring cultural relic structures is needed. Solutions have become an urgent need in the industry. Summary of the Invention

[0003] This invention overcomes the problems of subjective assumptions, inconsistent standards, and low deduction efficiency in existing technologies, and provides an AI-based proportional self- The method and system for the perfect and mature automatic restoration of cultural relics, derived from the deduction, quantifies the facial restoration threshold and combines symmetry constraints with... Using historical documents as a reference, we can achieve a scientific reconstruction of the structure that can be uniquely deduced, while preserving the incomplete parts that cannot be uniquely determined, thus ensuring... The authenticity, rigor, and standardization of cultural relic restoration.

[0004] The technical solution of this invention is as follows: (I) An automatic method for restoring cultural relics to their perfect mature state based on AI-based proportional self-consistent extrapolation Establishing a coordinate system for cultural relics themselves: based on the perfect, mature state of cultural relics throughout their life cycle, characterized by structural stability, proportional consistency, and functional integrity. Accuracy is achieved by constructing a four-dimensional native coordinate system that includes proportional dimensions, material properties, process logic, and mechanical constraints.

[0005] Data Acquisition and Feature Extraction: 3D point cloud, image, and material data of cultural relics are acquired through 3D scanning; AI automatically extracts unique features. A defined structural invariant characteristic.

[0006] Acquired disturbance removal: AI identifies and removes acquired disturbances such as deformation, damage, corrosion, and improper repairs made in the past.

[0007] AI-based proportional self-consistent continuity deduction: Based on residual structural features, a proportional self-consistent algorithm is used to complete the structural continuity from local to global. Through continuous deduction, a complete and proportionally accurate model of the cultural relic's form is generated.

[0008] Facial structure restoration specific rules: (1) Conditions for restoration: When the remaining facial area of ​​the cultural relic is ≥1 / 4 of the complete facial structure, and includes the facial midline or one side of the face. When shaping the facial contour, the complete facial structure can be derived and restored based on the left-right symmetry constraint. (2) Conditions under which restoration is prohibited: Restoration will not be performed when the remaining facial area is less than 1 / 4, or when it does not include the central axis or has no complete contour anchor points. Facial contour restoration, preserving the incomplete state; (3) Document-assisted reconstruction: For statues with clear identity references, their form descriptions in contemporary historical documents can be used as a reference. Quantifiable features such as eyebrow shape, eye shape, facial contour, and beard shape, when referenced to the proportional paradigm of similar and contemporary standard statues, are analyzed. (3) Auxiliary restoration under the constraints of symmetry and proportional continuity; (4) Exclusion cases: when there is no clear identity, no contemporary form standard, or the literature is only a literary exaggeration, the restoration shall not be based on the literature. Facial restoration was performed.

[0009] Triple verification and optimization: The model undergoes triple verification for structural consistency, morphological uniqueness, and physical stability, and is iteratively optimized to achieve uniqueness. A unified understanding.

[0010] Output: Output a fully functional, mature digital model of the cultural relic, restoration parameters, and a restoration plan; for items where a unique model cannot be determined... It retains its unique artistic characteristics while preserving its original, incomplete state.

[0011] (II) An Automatic Restoration System for Perfectly Mature Cultural Relics Based on AI-Based Proportional Self-Consistent Deduction It includes a data acquisition module, a coordinate system construction module, an AI data processing module, and a proportional self-consistent inference module, which are connected sequentially. The module includes a block-based facial constraint restoration module, a triple verification module, and a results output module. The data acquisition module is used to acquire 3D point cloud, image, and material data of cultural relics; The coordinate system construction module is used to establish a native coordinate system with the perfect and mature state of the cultural relic's own structure as the core; The AI ​​data processing module is used to extract structural invariants, remove acquired disturbances, and filter out unfounded artistic features. The proportional self-consistent deduction module is used to realize the structural continuity deduction from local to global; The facial constraint restoration module is used to perform facial structure restoration based on 1 / 4 threshold judgment, symmetry constraints, and document form assistance; The triple verification module is used to verify and optimize the restored model; The output module is used to output the structural restoration model and repair plan, while retaining the incomplete feature areas without basis. Beneficial effects

[0012] Rigorous and verifiable quantitative measures: The minimum threshold for recovery is clearly defined as 1 / 4 of the face, ensuring standardized, repeatable, and verifiable results. Structural restoration is scientifically unique: relying on proportional consistency and symmetry constraints, it only restores the uniquely identifiable parts, without subjective creation; The use of literature is reasonable and compliant: only quantifiable descriptions of form are used to assist in the reconstruction, literary exaggeration is excluded, and the norms of cultural relic research are met. Clear boundaries and unambiguous bottom lines: No supplementation will be made if the amount is less than 1 / 4 or if there are no anchor points or evidence; the principles of minimal intervention and authenticity must be strictly followed. Highly efficient, precise, and automated: AI replaces human labor in completing massive simulations and verifications, significantly improving efficiency and accuracy; It has strong universality and wide applicability: it is suitable for the digital preservation and restoration of various damaged cultural relics such as statues, stone carvings, bronzes, ceramics, and woodenware. Example

[0013] Select a fragmented stone carving with a clear identity; approximately one-third of the left side of the face remains, including the area from the brow to the jaw. His outline and central axis, according to contemporary literature, indicate that he had "a square face, broad jaw, silkworm eyebrows, phoenix eyes, and three long strands of beard."

[0014] Establish the original structural coordinate system of the statue and determine the proportions, mechanics, and technological standards. 3D scanning acquires residual facial point cloud and image data, and AI extracts structural invariant features; Eliminate external disturbances such as damage and weathering; If the remaining facial features are ≥1 / 4 and contain effective anchor points, the restoration conditions are met; AI restores the complete facial contour based on symmetry constraints, and combines the description of the form in the literature with the constraints of the contemporary iconographic paradigm on the proportion of facial features. After triple verification and optimization, a complete statue model with a unique structure and rigorous proportions was generated. The output includes a digital model and a restoration plan, and the restoration process is free from subjective fabrication. Example

[0015] One statue only has fragments of a cheek remaining, accounting for less than 1 / 8 of the face, with no central axis and no complete outline anchor points; although some text exists. The system determines that the restoration conditions are not met, the facial area remains incomplete, and no further reconstruction or completion is performed.

Claims

1. A method for automatically restoring cultural relics to a perfect, mature state based on AI-based proportional self-consistent extrapolation, characterized in that, Includes the following steps: Establish a native coordinate system centered on the perfect, mature state of cultural relics, characterized by structural stability, proportional consistency, and complete functionality; Collect 3D point cloud, image and material data of cultural relics, and use AI to extract uniquely identifiable structural invariant features of the remaining parts; AI identifies and removes acquired disturbances such as deformation, damage, and corrosion, restoring the original structural form of cultural relics; Based on the proportional self-consistency algorithm, the structural continuity of cultural relics from local to overall structure is deduced, and a structurally complete morphological model is generated by fitting. Quantitative constraints are applied to the restoration of the face of cultural relics: when the remaining facial features represent ≥1 / 4 of the complete facial structure and include the central axis or a unilateral contour. The facial features are reconstructed based on symmetry constraints, and the reconstruction can be aided by combining descriptions of form from historical documents of the same period with the proportional paradigms of similar statues; when If less than 1 / 4 of the face remains or there are no effective anchor points, facial restoration will not be performed. The model is subjected to triple verification of structural self-consistency, morphological uniqueness, and physical stability, and then iteratively optimized. Output a fully functional digital model of the cultural relic's structure and a restoration plan, while preserving any incomplete features that cannot be uniquely identified.

2. The method according to claim 1, characterized in that, The structural invariant features include scale, texture direction, Material properties, craftsmanship marks, and mechanical constraints do not include personalized artistic features without a unique basis.

3. The method according to claim 1, characterized in that, The document's format description includes eyebrow shape, eye shape, facial contour, The description includes quantifiable physical features such as the shape of the beard, and does not contain literary exaggerations.

4. The method according to claim 1, characterized in that, The triple verification must simultaneously satisfy: no structural contradictions, and restoration. Its form has a unique optimal solution, and its physical properties conform to the laws of the original materials of cultural relics.

5. An automatic restoration system for perfectly mature cultural relics based on AI-based proportional self-consistent extrapolation, characterized in that, Including sequential data The connected modules include data acquisition, coordinate system construction, AI data processing, proportional self-consistent deduction, and facial constraints. Restoration module, triple verification module, and results output module; The facial constraint restoration module is used to perform facial structure restoration based on 1 / 4 threshold determination, symmetry constraints, and document format assistance. Original.

6. The system according to claim 5, characterized in that, The system achieves fully automated structure restoration without generating any data. Artistic features that cannot be uniquely defined.

7. The system according to claim 5, characterized in that, The proportional self-consistent deduction module can uniquely simulate based on local structural features. The overall structural form should be integrated to ensure continuous restoration proportions, structural self-consistency, and mechanical rationality.