Self-Position Estimation Using Landmark Motion and Terrain Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional SLAM technologies face challenges in accurately localizing and mapping in environments with frequently moving objects, such as personal belongings, which are not effectively utilized due to low calculated influence, leading to decreased accuracy and instability in responding to changes in the surrounding environment.
Innovation Solution
An information processing device that acquires landmark and terrain data, estimates a range for self-position based on a landmark map, and further refines the self-position using a terrain data map, incorporating landmark movement ranges to stabilize localization and mapping processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional SLAM uses landmarks with low calculated influence (including frequently moving objects), then the system can maintain simplicity in landmark selection, but the localization accuracy decreases because useful moving landmarks are discarded
Solution Approach 1:
The patent applies dynamics by making the landmark selection process adaptive rather than static. The system dynamically adjusts which objects are selected as landmarks based on their movement characteristics and utility for localization. Frequently moving objects that were previously discarded are now incorporated as landmarks when they provide useful localization information, allowing the system to adapt to environmental changes while maintaining accuracy.
Solution Approach 2:
The patent changes the parameters used for landmark selection from purely static criteria to include dynamic movement patterns. By incorporating movement range information and calculating utility based on both position and movement characteristics, the system transforms the landmark selection process to identify useful moving landmarks that were previously excluded by conventional static selection methods.
2Adaptability or versatility
If the system discards frequently moving objects as landmarks, then the landmark selection process remains simple, but the system cannot respond to changes in the surrounding environment
Solution Approach 1:
The system implements dynamic landmark selection that adapts to environmental changes by monitoring object movement patterns. Rather than using fixed selection criteria, the system continuously evaluates which moving objects serve as useful landmarks based on their movement characteristics and localization utility, enabling responsive adaptation to environmental dynamics.
Solution Approach 2:
The patent incorporates feedback mechanisms where the system evaluates the utility of moving objects as landmarks based on their movement patterns and localization performance. This feedback loop allows the system to identify and select useful moving landmarks while filtering out truly problematic objects, balancing adaptability with computational efficiency.
3Measurement precision
If conventional SLAM calculates low influence for frequently moving landmarks, then the processing remains computationally efficient, but the localization accuracy decreases
Solution Approach 1:
The patent transforms the computational approach by changing the parameters evaluated for landmark selection. Instead of using simple static criteria that discard moving objects, the system evaluates movement patterns, ranges, and utility metrics to identify useful moving landmarks. This parameter transformation enables accurate localization using moving objects without requiring excessive computational resources for complex tracking and analysis.
Data Source
AI summary
An information processing device according to the present disclosure includes an acquisition unit that acquires an arrangement of a landmark and surrounding terrain data, a first estimation unit that estimates a range in which a self-position is located on a basis of a landmark map that is a map indicating the arrangement of the landmark, and a second estimation unit that estimates the self-position from the range in which the self-position is located on a basis of a terrain data map that is a map indicating the surrounding terrain data.