Zero-Watermarking BIM Data via Vertical Stability Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing zero-watermarking algorithms for BIM data face challenges in robustness against primitive attacks, particularly due to the fine monomer nature and low redundancy of BIM data, which affects the extraction of invariants and construction of zero-watermarking based on global geometric features, and the sensitivity of local feature-based algorithms to primitive additions or deletions.
Innovation Solution
A zero-watermarking method that clusters primitives based on vertical stability, using norm skewness measurement to construct a mapping relationship between watermarking bits and spatial positions, and performs an XOR operation to generate a robust zero-watermarking sequence, enhancing resistance to translation, rotation, and primitive attacks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If zero-watermarking is based on global geometric features, then watermarking robustness is improved, but it requires high redundancy in the original data which BIM data lacks
Solution Approach 1:
The patent segments the BIM model into multiple view projections (front view, back view, left view, right view, top view, bottom view) and extracts geometric features from each projection separately. This segmentation allows the method to work with the low-redundancy characteristics of BIM data by distributing the watermarking process across multiple views rather than requiring global redundancy.
Solution Approach 2:
The patent transforms the 3D BIM model into 2D projection views and performs watermarking in this different dimensional space. By converting from 3D to 2D projections and using 2D image processing techniques (DCT transformation, quantization), the method adapts global feature-based watermarking to BIM data structures.
2Adaptability or versatility
If zero-watermarking is based on local geometric features, then flexibility and attack resistance are improved, but sensitivity to primitive additions or deletions increases
Solution Approach 1:
The patent merges multiple local view projections into a comprehensive watermarking representation. By combining features from front, back, left, right, top, and bottom views, the method creates a more robust watermark that is less sensitive to local primitive modifications, as the damage would need to affect multiple views simultaneously.
Solution Approach 2:
The patent uses DCT (Discrete Cosine Transformation) which is a universal transform applicable to both 2D images and the projected views of 3D BIM models. This multi-functional approach allows the same mathematical framework to handle different view projections and provides consistent watermarking performance across various attack scenarios.
3Reliability
If traditional watermarking embeds copyright information into original data, then copyright protection is achieved, but accuracy loss or errors in BIM data occur
Solution Approach 1:
The patent introduces DCT coefficients as an intermediary domain for watermarking. Instead of directly modifying the original BIM geometric data, the method transforms view projections into the frequency domain, embeds watermark information in the DCT coefficient space, and then reconstructs the data. This intermediary approach protects copyright while minimizing impact on the precision of the original BIM data.
Solution Approach 2:
The patent creates a transformed copy of the BIM view projections in the frequency domain (DCT domain) and performs watermarking on this copy rather than the original spatial data. The watermarked copy is then transformed back, ensuring that the original BIM data precision is preserved while copyright protection is embedded in the transformed representation.
Data Source
AI summary
A zero-watermarking method and device for BIM data, and a medium are provided, which relate to the field of watermarking information security technologies. Aiming at existing zero-watermarking method for the BIM data cannot resist primitive attacks, vertical stability of a model is used to construct a mapping relationship between primitive clusters and watermarking bits, calculate norms of primitives in each primitive cluster of the primitive clusters, take positivity and negativity of norm skewness measurement of the primitives as eigenvalues to construct a binary sequence, and performs an XOR process on the binary sequence and an original watermarking sequence to construct the zero-watermarking for the BIM data. Experimental results indicate that the zero-watermarking method has uniqueness, robustness and security.