3D Collision Judgment Using Point Clouds and Voxels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current techniques for judging collisions between structures in a charged particle beam apparatus's sample chamber are inadequate as they rely on two-dimensional representations, failing to account for depth and often inaccurately determine collisions due to predefined shapes, leading to potential collisions going undetected before actual movement.
Innovation Solution
A collision judgment apparatus that uses three-dimensional models, represented by point group information and voxel group information, to accurately determine overlaps between dynamic and static objects within the sample chamber, considering movement paths, tilts, and rotational angles, thereby preventing collisions before actual movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If two-dimensional representations are used for collision judgment, then the judgment process is simple, but the accuracy of collision detection is insufficient
Solution Approach 1:
The patent transitions from two-dimensional collision judgment to three-dimensional collision judgment by introducing depth information through the Z-axis. The sample holder position is represented by coordinates (x, y, z) and the objective lens position by (x0, y0, z0), enabling accurate detection of spatial relationships including depth, thereby resolving the contradiction between simplicity and accuracy.
2Productivity
If predefined shapes are used for collision judgment, then the judgment process is efficient, but false collision detection occurs
Solution Approach 1:
The patent changes the parameters used for collision judgment from fixed predefined shapes to dynamic coordinate-based parameters. The actual positions of the sample holder (x, y, z) and objective lens (x0, y0, z0) are used, along with actual dimensions such as sample height h, enabling accurate collision detection that adapts to real configurations rather than relying on predetermined shapes.
3Measurement precision
If three-dimensional models are used for collision judgment, then the accuracy of collision detection is improved, but the time required for judgment increases
Solution Approach 1:
The patent segments the collision judgment process into distinct computational steps: calculating the distance d between sample holder and objective lens using the three-dimensional coordinates, comparing this distance against the sum of their radii (r + r0), and making a collision determination based on this comparison. This segmented approach enables efficient three-dimensional collision judgment without excessive computational time.
Data Source
AI summary
Based on a three-dimensional model of a dynamic object having a position which changes, point group information which represents, with a group of points, a three-dimensional shape of the dynamic object is generated. Based on a three-dimensional model of a static object having a position which does not change, point group information representing, with a group of points, a three-dimensional shape of the static object is generated. Based on the point group information of the static object, voxel group information which represents, with a group of voxels, the three-dimensional shape of the static object, and which is formed into a database is generated. Presence or absence of overlap between the dynamic object and the static object is judged by collating voxel group information representing the static object which is present on a movement path of the dynamic object, and point group information representing the dynamic object.