Virtual Gimbal Registration for 3D Depth Scan Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 3D depth scanning technologies require computationally intensive methods for point cloud registration, which are not readily implementable on consumer devices like smartphones due to large data sets and sensitivity to initial conditions.
Innovation Solution
The use of virtual gimbal information, derived from orthogonal axes of features in range images, to define a coordinate system for efficient registration of 3D depth scans, allowing for real-time alignment without iterative processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If iterative methods such as Iterative Closest Point or Polygonal Mesh are used for point cloud registration, then alignment precision can be achieved, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent extracts and utilizes gimbal information from the imaging device as a separate, independent data source for registration. By separating the gimbal orientation data from the point cloud data, the system can perform registration using only gimbal information, avoiding the need to process large point cloud datasets while maintaining alignment precision.
Solution Approach 2:
The patent introduces gimbal information as an intermediary element that mediates between the imaging device and the point cloud registration process. The gimbal orientation data serves as a bridge that enables direct calculation of coordinate transformations without requiring iterative point cloud matching, thus reducing computational complexity while preserving alignment accuracy.
2Measurement precision
If iterative registration methods are used, then accurate alignment can be achieved, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary action by capturing and storing gimbal orientation information at the time of image acquisition. This pre-captured gimbal data is then directly used for registration calculations, eliminating the need for time-consuming iterative processing during the registration phase and enabling rapid alignment.
Solution Approach 2:
The patent replaces the mechanical iterative optimization process with a direct mathematical calculation based on gimbal orientation data. Instead of using iterative algorithms that repeatedly adjust point cloud positions, the system substitutes this with a closed-form solution that calculates the coordinate transformation directly from gimbal angles, significantly reducing processing time.
3Adaptability or versatility
If point cloud registration is performed on consumer devices, then mobile 3D scanning becomes accessible, but computational limitations of consumer devices prevent implementation of intensive algorithms
Solution Approach 1:
The patent extracts only the essential gimbal orientation parameters from the imaging device, separating this lightweight data from the heavy point cloud data. This extracted gimbal information can be processed on consumer devices with limited computational resources, making mobile 3D scanning accessible while avoiding the need to run intensive algorithms on these devices.
Solution Approach 2:
The patent changes the parameters used for registration from full point cloud coordinates to simplified gimbal orientation angles. This parameter transformation reduces the data dimensionality and computational requirements, enabling registration to be performed on consumer devices that lack the processing power for traditional point cloud algorithms.
4Quantity of substance
If large point cloud datasets are processed for registration, then comprehensive scene coverage is achieved, but memory and computational requirements exceed consumer device capabilities
Solution Approach 1:
The patent extracts and utilizes only the gimbal orientation data from the imaging device, completely separating this lightweight metadata from the large point cloud datasets. This extraction approach maintains data completeness for registration purposes while reducing memory requirements to minimal levels that consumer devices can easily handle.
Solution Approach 2:
The patent segments the registration process into two independent parts: capturing gimbal orientation information during data acquisition, and using this segmented information for registration calculations. This segmentation allows the system to maintain comprehensive scene coverage without requiring the device to load and process large point cloud datasets into memory.
Data Source
AI summary
Systems and methods configured to generate virtual gimbal information for range images produced from 3D depth scans are described. In operation according to embodiments, known and advantageous spatial geometries of features of a scanned volume are exploited to generate virtual gimbal information for a pose. The virtual gimbal information of embodiments may be used to align a range image of the pose with one or more other range images for the scanned volume, such as for combining the range images for use in indoor mapping, gesture recognition, object scanning, etc. Implementations of range image registration using virtual gimbal information provide a realtime one shot direct pose estimator by detecting and estimating the normal vectors for surfaces of features between successive scans which effectively imparts a coordinate system for each scan with an orthogonal set of gimbal axes and defines the relative camera attitude.