Camera-Based Positioning Using Self-Addressing Sources for Dead Reckoning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current navigation systems, such as personal navigation devices (PNDs), face challenges in providing accurate position estimates when Global Navigation Satellite System (GNSS) signals are blocked or weak, especially in urban environments or indoors, leading to errors in dead reckoning due to accumulated inertial sensor measurement errors.
Innovation Solution
A method and apparatus that utilize a camera-based system to detect self-addressing sources (SAS) and compute vectors to recalibrate the dead-reckoning navigation system, allowing for precise position determination without the need for extensive a priori mapping, by using image processing to determine vectors and correct position estimates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dead reckoning is used for position estimation, then navigation can continue without GNSS signals, but position error accumulates over time
Solution Approach 1:
The system uses visual odometry to continuously monitor and correct the dead reckoning position estimates. By detecting features in the environment and tracking their movement across frames, the system provides feedback that identifies and corrects accumulated errors in the inertial navigation system's position estimates, thereby maintaining both continuity and accuracy.
Solution Approach 2:
The patent replaces the purely mechanical/inertial dead reckoning system with a hybrid system that incorporates visual processing. The camera-based visual odometry system complements and corrects the inertial measurements, substituting the error-prone mechanical accumulation with an optical measurement approach that can detect and correct position drift.
2Reliability
If cellular-based position location techniques are used, then position can be determined without GNSS, but position resolution is low and uncertainty is high
Solution Approach 1:
The system replaces cellular-based radio positioning with camera-based visual odometry. Instead of relying on radio wave propagation characteristics that yield low-resolution position data, the system uses optical imaging and feature tracking to achieve high-precision position estimation, substituting the radio-based measurement mechanism with an optical one.
3Measurement precision
If a priori mapping and triangulation are used for WiFi-based positioning, then position can be estimated, but extensive calibration and measurements are required
Solution Approach 1:
The system performs self-calibration by automatically detecting and tracking features in the environment without requiring pre-established maps or extensive manual calibration. The visual odometry algorithm adapts to the environment as the device moves, automatically establishing the spatial relationships between features and the device position, thereby eliminating the need for complex a priori mapping and calibration procedures.
Data Source
AI summary
An apparatus and method for estimating a position of a mobile device based on one or more images is disclosed. Positioning information is derived from an image containing an SAS (self-addressing source such as a QR code), thereby setting a first vector V1 and an SAS pose. The image is also used to determine a displacement between the mobile device and the self-addressing source, thereby setting a second vector V2. Using the first vector V1, the SAS pose and the second vector V2, the mobile device may estimate its position with high accuracy and may also recalibrate dead-reckoning navigation and a gyrometer.