3D Dental Image Registration via Overlap Quality Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for registering individual three-dimensional optical recordings of dental objects often result in incorrect registrations due to insufficient overlapping areas, recording errors, or disturbing objects, leading to distorted images and the need for repeated measurements.
Innovation Solution
A method where each individual recording is automatically checked for registration conditions, and if met, the recordings are combined into clusters, with new sequences started if conditions are not met, allowing for error-free and efficient formation of an overall recording by ensuring sufficient overlapping areas and recording quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If individual three-dimensional optical recordings are registered based on overlapping areas, then an overall recording of the dental object is formed, but incorrect registration may occur due to insufficient overlapping areas, recording errors, or disturbing objects
Solution Approach 1:
The system automatically checks the quality of overlapping areas between consecutive recordings and provides feedback to determine whether to continue or abort the recording sequence. This quality control mechanism ensures that only recordings meeting predetermined quality criteria are used for registration, thereby maintaining high measurement precision while preventing erroneous registrations.
Solution Approach 2:
The system performs preliminary quality assessment of overlapping areas before final registration. By checking whether overlapping areas meet quality criteria in advance, the system prevents incorrect registrations from occurring in the first place, rather than detecting and correcting them after registration has taken place.
2Reliability
If the entire survey is repeated when registration is found to be incorrect, then error-free measurement can be achieved, but measurement time increases significantly
Solution Approach 1:
The system performs preliminary quality checks of overlapping areas during the measurement process itself. By detecting potential registration errors before they compromise the overall measurement, the system avoids the need to repeat entire surveys, thereby maintaining high reliability while minimizing time loss.
Solution Approach 2:
When a recording sequence is aborted due to poor overlapping area quality, the system skips continuing that problematic sequence and immediately starts a new one. This allows the system to quickly move past erroneous recordings rather than wasting time attempting to correct or reprocess failed registrations.
3Measurement precision
If automatic quality checking and sequence aborting is implemented, then registration accuracy is improved, but system complexity increases
Solution Approach 1:
The system performs automatic quality checking and sequence aborting without requiring manual intervention. The computer automatically evaluates overlapping area quality, determines whether to continue or abort recording sequences, and manages the formation of clusters. This self-service capability improves registration accuracy while keeping operational complexity low, as the system handles all quality control decisions autonomously.
Data Source
Figure 1
AI summary
The invention relates to a method for recording individual three-dimensional optical images (2) to form a global image of a dental object (1) which is to be measured. After each individual image (2) is taken by means of a dental camera (3), a computer (8) automatically checks whether an overlapping area (5, 12) between the images which are to be combined meet the recording requirements determined for a correct recording. If the overlapping area (5) meets the recording requirements, the recording is carried out between the images (6, 7) to be combined and a first image sequence (9) is set in motion. Said images of the first image sequence (9) are subsequently combined to form a first cluster (23). If the overlapping area (12) of an image (13) does not meet the recording requirements, the first image sequence (9) is interrupted, an additional second image sequence (14) is automatically started with said image (13). Subsequently, the images of the second image sequence (14) are combined to form a second cluster (24).