Image-Based Navigation Using Fundamental Matrix Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current navigation systems, such as GPS and INS, face limitations including interference, indoor and urban area obstructions, and accuracy issues, while image-based navigation methods are slow and inaccurate, necessitating an improved image-based navigation method that provides accurate position and orientation.
Innovation Solution
An image-based navigation system using an optimized fundamental matrix calculates rotation and translation from correlated point features in images, employing single value decomposition (SVD) and the LevenBerg-Marquard algorithm to determine relative camera movement, correcting for radial distortion and other image effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If image-based navigation techniques are used, then navigation can be provided without GPS satellites or inertial sensors, but the system becomes slow and inaccurate compared to GPS and INS
Solution Approach 1:
The patent transforms the navigation problem by changing the mathematical parameters used for calculation. It employs an optimized fundamental matrix with simultaneous iteration of rotation and translation parameters, rather than traditional sequential methods. This parameter optimization enables the system to achieve GPS/INS-level accuracy by refining the mathematical model of image-based navigation.
Solution Approach 2:
The patent implements an iterative feedback mechanism where the fundamental matrix is continuously refined through simultaneous calculation and iteration of rotation and translation. The system uses feedback from correlated point features across multiple images to progressively improve position and orientation estimates, achieving convergence to high-precision navigation data.
2Measurement precision
If traditional fundamental matrix methods are used, then rotation and translation can be calculated from image points, but the calculation is slow and not optimized for real-time navigation
Solution Approach 1:
The patent performs preliminary optimization of the fundamental matrix before final navigation calculation. By pre-processing image correlations and establishing an optimized fundamental matrix relationship, the system prepares the mathematical framework in advance, enabling faster real-time iteration and calculation of rotation and translation parameters during actual navigation operations.
Solution Approach 2:
The patent changes the calculation approach by simultaneously iterating rotation and translation parameters rather than calculating them sequentially. This parameter change optimizes the computational path, reducing the number of iterations needed and improving real-time navigation calculation speed while maintaining high accuracy.
3Reliability
If navigation systems rely on GPS satellites, then position and orientation can be provided globally, but the system fails indoors and in urban areas with obstructed line of sight
Solution Approach 1:
The patent replaces the satellite-based mechanical navigation system with an image-based optical system. Instead of relying on GPS satellites and radio signals that require line of sight, the system uses cameras to capture visual features and compute navigation data through image correlation and fundamental matrix calculations, enabling operation in environments where GPS is unavailable.
Solution Approach 2:
The patent creates a universal navigation system that can operate across diverse environments including outdoor areas, indoor spaces, and urban canyons. The image-based fundamental matrix method provides multi-functional capability to handle various lighting conditions, scene types, and spatial configurations, making the system adaptable to any environment with visual features.
Data Source
AI summary
A method and apparatus for determining position and orientation enabling navigation of an object using image data from at least a first and a second 2D image from at least one camera mounted on said object. The method comprises the steps of: correcting images from one or several cameras and from at least a first and a second 2D image for their respective radial distortion and other measurable effects which result in poor image precision; matching 2D image items in and between at least a first and second 2D image; calculating a fundamental matrix by using correlated image points from at least a first and a second 2D image; calculating and extracting estimated first rotation and translation values from the fundamental matrix using single value decomposition (SVD) based on information from at least a first and a second 2D image; iterating more accurate final rotation and translation values by using the LevenBerg-Marquard algorithm and determining the position and orientation of said object.