Estimation Apparatus for Moving Object Position and Posture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for estimating the position and posture of a second moving object in the periphery of a first moving object, such as vehicles, face challenges in accuracy, especially when the density of measurement points is low, leading to reduced estimation precision.
Innovation Solution
An estimation apparatus equipped with a processor that acquires measurement points using a distance sensor, classifies candidate points based on past and current measurements, and calculates an evaluation value using likelihood methods to estimate the position and posture of the second moving object, even at increased distances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If three-dimensional measurement is performed to estimate position and posture of a second moving object, then estimation capability is achieved, but measurement precision deteriorates when measurement point density is low
Solution Approach 1:
The patent introduces an intermediary processing step that classifies measurement points into candidate points based on spatial relationships and movement vectors. This classification acts as a mediator that extracts meaningful information from sparse measurement data, enabling accurate estimation even when measurement point density is low. The intermediary process transforms raw measurement data into structured candidate information that can be reliably used for position and posture estimation.
2Area of stationary object
If measurement distance is increased to expand detection range, then coverage area is improved, but measurement precision deteriorates due to sparse measurement points
Solution Approach 1:
The patent transitions from two-dimensional measurement plane analysis to three-dimensional spatial relationship analysis. By considering measurement points in three-dimensional space and analyzing their spatial relationships, movement vectors, and positional changes over time, the system maintains estimation precision even when measurement points are sparse due to increased detection distance. This dimensional transformation enables effective utilization of limited measurement data.
3Measurement precision
If classification of candidate points is performed to improve estimation accuracy, then estimation precision is improved, but device complexity increases
Solution Approach 1:
The patent segments the measurement point classification process into distinct functional stages: identifying candidate points based on spatial relationships, calculating movement vectors, and determining position and posture. This segmentation allows each processing stage to be optimized independently and enables parallel processing where applicable, reducing overall computational complexity while maintaining high estimation precision through systematic classification.
Data Source
AI summary
According to an embodiment, an estimation apparatus includes a memory and a processor. The processor acquires a first measurement point groups obtained by measuring a periphery of a first moving object. The processor estimates a position and posture of the first moving object. The processor classifies first measurement points serving as candidates of measurement points on a second moving object in the newest first measurement point group as candidate points. The processor acquires second moving object information from the second moving object. The processor calculates an evaluation value using a first likelihood defined according to a position relationship between an orientation of a region specified from the second moving object information and the candidate points. The processor estimates a position and posture of the second moving object based on the evaluation value.