Self-Position Estimation with Adaptive Feature-Point Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing self-position estimation methods in autonomous robots face accuracy deterioration due to increased processing load or reduced feature point usage, leading to inefficient visual simultaneous localization and mapping (VSLAM) processes.
Innovation Solution
An information processing apparatus that dynamically adjusts the number of feature points detected from images based on the number of corresponding feature points, ensuring accurate self-position estimation while reducing processing load by optimizing feature point detection and estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the number of feature points to be extracted from the input image is increased to maintain self-position estimation accuracy, then the accuracy of self-position estimation is improved, but the processing time required to detect the position and posture of the imaging device increases
Solution Approach 1:
The patent applies dynamics by making the number of extracted feature points variable rather than fixed. The detection unit dynamically adjusts the number of feature points to extract based on real-time processing time requirements, allowing the system to adapt between accuracy and speed needs during operation
Solution Approach 2:
The patent changes the parameter of feature point quantity based on processing time conditions. When processing time is sufficient, more feature points are extracted for higher accuracy; when processing time is constrained, fewer feature points are extracted to meet timing requirements
2Quantity of substance
If the number of images to be thinned out is increased to reduce the number of images to be stored, then the storage requirements are reduced, but the accuracy of self-position estimation deteriorates
Solution Approach 1:
The patent applies local quality by selectively retaining images based on their contribution to feature point correspondence. Instead of uniform thinning, the system identifies and preserves specific images that contain critical feature points needed for accurate self-position estimation, while thinning out less important images
3Adaptability or versatility
If the processing load of processes other than position and posture detection is increased, then the functionality of the autonomous movement apparatus is enhanced, but the processing time required to detect the position and posture of the imaging device increases
Solution Approach 1:
The patent applies partial action by extracting only the necessary number of feature points required for position and posture detection, rather than processing all possible feature points. This allows other processes to share computational resources without excessively impacting the detection timing
Data Source
AI summary
An information processing apparatus according to the present disclosure includes: a detection unit configured to detect a plurality of new feature points from a first image; a specification unit configured to specify, among the plurality of new feature points, a corresponding feature point corresponding to a known feature point associated with a three-dimensional position included in at least one management image used for generating an environment map; and an estimation unit configured to estimate a position and a posture of an imaging device that has captured the first image, by using the corresponding feature point, in which the detection unit changes the number of new feature points to be detected from a target image that is a target for estimating the position and the posture of the imaging device according to the number of corresponding feature points.