Multi-Sensor Pose-Estimate System Using Virtual Image Aggregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current pose-estimation systems face challenges in accurately determining the three-dimensional position and orientation of a target object from two-dimensional image data, particularly when using non-overlapping image sensors, as they often rely on stereo vision and are prone to errors related to baseline length and collinearity.
Innovation Solution
A multi-sensor pose-estimate system that employs a vision system with multiple image sensors capturing images on partially non-overlapping planes, using a pose-estimation processor to aggregate image data and implement a least squares summation function to minimize error vectors, incorporating baseline spatial and orientation data to generate a pose estimate in three-dimensional space without relying on stereo vision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If stereo vision is used for pose estimation, then depth information can be obtained, but baseline length errors and collinearity errors occur
Solution Approach 1:
The system divides the imaging task into multiple independent image sensors positioned at different locations, each capturing images from different perspectives. Instead of using a single stereo pair with baseline errors, multiple sensors segment the measurement task to eliminate collinearity errors by ensuring non-collinear feature points across multiple images.
Solution Approach 2:
The system transitions from two-dimensional stereo vision to multi-dimensional pose estimation by incorporating multiple image sensors positioned in three-dimensional space. This allows the system to determine pose from multiple angular perspectives simultaneously, eliminating the baseline length error constraint of traditional stereo vision.
2Area of stationary object
If multiple image sensors with non-overlapping planes are used, then coverage area increases, but image aggregation complexity increases
Solution Approach 1:
The system merges multiple images from sensors with non-overlapping image planes by identifying and matching common feature points across all images. The pose estimation processor aggregates these diverse images into a unified pose estimate, transforming complexity into a solvable feature-matching problem.
Solution Approach 2:
The pose estimation processor serves multiple functions: it identifies feature points across different image planes, performs coordinate transformations, calculates pose parameters, and validates results. This multi-functional approach handles the complexity of non-overlapping image aggregation through a single integrated processing system.
Data Source
AI summary
One example includes a pose-estimate system. The system includes a vision system that includes a plurality of image sensors. Each of the plurality of image sensors captures an image with respect to a respective image plane. At least one of the respective images includes a feature of a target object. The system also includes a pose-estimation processor that aggregates the images corresponding to the respective image planes onto a virtual image plane to generate a pose estimate of the feature of the target object in three-dimensional space.