Low-Latency SLAM Using Sparse Short-Window 3D Position Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Standard SLAM methods are slow, power-consuming, and cause latency and motion sickness due to delayed object positioning and orientation updates, especially with fast-moving objects, leading to incomplete scans and distorted views.
Innovation Solution
A method for simultaneous localization and mapping (SLAM) that identifies a sparse set of object locations within a shorter time window, transforming these locations to a best-fit dense set of 3D positions using triangulation, allowing low-latency, efficient mapping and localization by supplementing sparse data with dense data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If full scene scanning is performed to locate objects, then positioning accuracy is improved, but processing time increases and power consumption increases
Solution Approach 1:
The patent segments the scanning process by dividing the scene into multiple regions and scanning only specific regions containing moving objects rather than performing full scene scanning. This selective region-based approach maintains positioning accuracy while reducing processing time and power consumption.
Solution Approach 2:
The patent applies partial action by scanning only the necessary portions of the scene (regions with moving objects) rather than the entire scene. This partial scanning strategy achieves sufficient positioning accuracy for fast-moving objects while significantly reducing processing time and energy consumption.
2Measurement precision
If full scene scanning is performed to locate objects, then positioning accuracy is improved, but power consumption increases
Solution Approach 1:
The patent segments the scanning process by dividing the scene into multiple regions and scanning only specific regions containing moving objects rather than performing full scene scanning. This selective region-based approach maintains positioning accuracy while reducing processing time and power consumption.
Solution Approach 2:
The patent applies partial action by scanning only the necessary portions of the scene (regions with moving objects) rather than the entire scene. This partial scanning strategy achieves sufficient positioning accuracy for fast-moving objects while significantly reducing processing time and energy consumption.
3Loss of information
If standard SLAM scanning is performed, then complete object information is obtained, but latency increases causing motion sickness
Solution Approach 1:
The patent applies preliminary action by performing fast scanning to obtain preliminary object position and motion information before conducting more detailed analysis. This preliminary fast scanning reduces latency and enables real-time tracking of fast-moving objects, preventing motion sickness while maintaining information completeness through subsequent refinement.
Solution Approach 2:
The patent implements dynamic scanning strategies that adapt to object motion characteristics. For fast-moving objects, the system uses shorter scanning intervals and focused region scanning to reduce latency, while maintaining complete object information through dynamic adjustment of scanning parameters based on detected motion patterns.
4Measurement precision
If scanning time is extended to capture fast-moving objects, then object tracking accuracy is improved, but motion distortion increases
Solution Approach 1:
The patent implements dynamic scanning strategies that adapt to object motion characteristics. For fast-moving objects, the system uses shorter scanning intervals and focused region scanning to reduce latency, while maintaining complete object information through dynamic adjustment of scanning parameters based on detected motion patterns.
Solution Approach 2:
The patent uses feedback mechanisms where object motion information detected in previous scanning cycles informs subsequent scanning operations. This feedback loop allows the system to track fast-moving objects accurately by adjusting scanning parameters based on measured motion, maintaining object representation stability while achieving high tracking accuracy.
Data Source
AI summary
The present invention relates to a method for sensing. The method comprises the step of identifying a second set of locations (2) of points of at least one object (3) in a scene (4) within a second time window (T2), with respect to a first reference point (28), such as preferably a first viewpoint. The method further comprises the step of converting said second set of locations (2) to a second set of 3D positions (6). The method further comprises the step of determining a transformation to said second set of 3D positions (6), such that said transformed second set of 3D positions (6) is a best fit to a first set of 3D positions (5), wherein said first set of 3D positions (5) is denser than said second set of 3D positions (6), and wherein said first set of 3D positions (5) is within a first time window (T1) prior to said second time window (T2), and wherein said second time window (T2) is shorter than said first time window (T1), preferably at least two times shorter.