Stereo Machine Vision for Natural Object Location
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Stereo machine vision systems face challenges in determining the locations of natural objects without predefined markers, as existing methods rely on disparity between camera sensors and struggle in non-controlled environments where attaching markers is inconvenient.
Innovation Solution
A system and method using a pair of optical devices with sensor arrays and a computing device to capture images and determine three-dimensional (x, y, z) coordinates of natural objects' surfaces by matching sensor readings across the devices, employing calibration tables to associate two-dimensional locations with three-dimensional positions, and generating surfaces and histograms to identify planes and orientations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If predefined markers are attached to target elements, then measurement precision is improved, but ease of operation deteriorates in non-controlled environments
Solution Approach 1:
The system uses natural features of objects themselves (edges, corners, surface textures) as target elements, eliminating the need for external markers. The objects serve their own purpose of being detected through their inherent geometric and textural properties, making the system applicable to non-controlled environments where marker attachment is impractical.
Solution Approach 2:
The invention extracts and utilizes naturally occurring geometric features (edges, corners, lines) from the target objects themselves, removing the requirement for artificial markers. This extraction of useful information from the objects' own structure enables operation in environments where markers cannot be attached.
2Device complexity
If disparity-based methods are used for natural objects, then device complexity is reduced, but measurement precision deteriorates
Solution Approach 1:
The system focuses on detecting specific local geometric features (edges, corners, lines) with high precision rather than attempting to analyze entire objects or surfaces. By concentrating computational resources on identifying and measuring these distinctive local features, the system achieves high measurement precision while maintaining relatively simple device architecture.
3Reliability
If marker-based systems are deployed, then reliability is improved, but adaptability deteriorates to non-controlled environments
Solution Approach 1:
The system is designed to detect multiple types of natural geometric features (edges, corners, lines, surface textures) across diverse objects and environments. This universal approach to feature detection enables the system to adapt to various non-controlled environments while maintaining reliable target identification, eliminating the need for environment-specific marker attachment procedures.
Data Source
AI summary
This document disclose stereo machine vision systems and methods for determining locations of surfaces of natural objects within the field of view of the stereo machine vision system whereby these natural objects are used as the target elements for the vision system.