Mass Estimation Using Modified Marching Cubes and Cube Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for processing data from 3D scanners, such as MRI, CT, and PET scanners, face challenges in efficiently estimating the mass of objects from density data, particularly in real-time applications and with limited computational resources, as existing algorithms like ray-casting and texture-based methods are CPU-intensive or lack detailed object information.
Innovation Solution
The implementation of a modified Marching Cubes algorithm that refines surface representations by segmenting objects, applying local movement vectors, and calculating mass contributions from cubes entirely within and intersecting with the surface, using density values to estimate mass through summation of volume contributions from sub-cubes within and intersecting with the surface.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If ray-casting or texture-based methods are used to process 3D scanner data, then detailed object information can be obtained, but the computational complexity increases and real-time processing becomes difficult
Solution Approach 1:
The patent divides the 3D scanned object into discrete cubic elements (voxels) that can be independently processed. Each cube is classified as either inside or outside the object surface, enabling parallel computation and reducing the complexity of mass estimation while maintaining accuracy through systematic volume integration.
Solution Approach 2:
The patent replaces traditional CPU-intensive ray-casting algorithms with a grid-based cubic segmentation approach that can be efficiently processed using parallel computing architectures. This substitution reduces computational complexity while preserving the ability to estimate mass from density data.
2Productivity
If traditional mass estimation methods are used, then computational resources can be minimized, but real-time processing capability is lost
Solution Approach 1:
By segmenting the object into discrete cubes, the patent enables parallel processing where multiple cubes can be evaluated simultaneously. This segmentation transforms a sequential computational problem into a parallelizable one, achieving real-time processing speeds while maintaining reasonable resource consumption through efficient algorithms.
Solution Approach 2:
The patent evaluates only the necessary cubic elements that contribute to mass estimation (those intersecting or containing the object surface), avoiding unnecessary computation on empty space. This partial action approach optimizes resource usage by focusing computational effort only where needed for accurate mass calculation.
3Measurement precision
If simple volume-based methods are used, then computational resources are conserved, but detailed object information and mass accuracy are lost
Solution Approach 1:
The patent applies different processing rules to different regions: cubes entirely inside the object use one calculation method, cubes entirely outside use another, and intersecting cubes use a third method involving surface intersection calculations. This local quality approach ensures high mass calculation accuracy by treating each region according to its specific geometric relationship with the object surface.
Solution Approach 2:
The patent introduces cubic elements as intermediary structures between the continuous 3D scanned data and the discrete mass calculation. These cubes serve as mediators that systematically bridge the gap between detailed surface geometry and volumetric mass estimation, preserving accuracy while enabling efficient computation.
4Manufacturing precision
If complex surface refinement algorithms are applied, then surface representation accuracy improves, but processing time increases
Solution Approach 1:
The patent performs preliminary surface segmentation and cube classification before mass calculation, organizing the data into a structured grid format. This preliminary action prepares the data in advance for efficient parallel processing during the actual mass estimation phase, reducing processing time while maintaining surface representation accuracy.
Solution Approach 2:
By segmenting both the surface representation and the volume into discrete cubes, the patent creates a unified grid-based framework that simultaneously achieves accurate surface representation and efficient mass calculation. The segmentation allows independent parallel processing of surface and volume attributes without increasing overall processing time.
Data Source
AI summary
The mass of an object may be estimated based on intersection points of a representation of a surface in an image space with cubes defining the image space, the surface representing a surface of an object. The representation may be, for example, based on marching cubes. The mass may be estimated by estimating a mass contribution of a first set of cubes contained entirely within the representation of the surface, estimating a mass contribution of a second set of cubes having intersection points with the representation of the surface, and summing the estimated mass contribution of the first set of cubes and the estimated mass contribution of the second set of cubes. The object may be segmented from other portions of an image prior to estimating the mass of the object.