Helmet Localization Using Inertial-Visual Screening
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current visual odometer localization methods fail to meet the accuracy requirements for dynamic objects in scenarios like firefighting and adventure due to hardware limitations, the need for pre-erected base stations, and the object's fast movement.
Innovation Solution
A localization method involving a wearable device like a helmet that extracts feature points from images, uses inertial information to screen these points, triangulates them to generate 3D map points, performs error loopback calibration, and determines positional points, enhancing accuracy for dynamic objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual odometer localization is used, then cost is reduced and image information is provided, but localization accuracy deteriorates for dynamic objects
Solution Approach 1:
The patent combines visual odometer localization with inertial navigation system (INS) to form a hybrid localization system. The visual system provides rich image information while the INS provides high-frequency motion data, and their integration through loose coupling achieves improved localization accuracy for dynamic objects without requiring complex tight coupling mechanisms
Solution Approach 2:
The patent introduces an intermediary processing mechanism that uses inertial information to screen and select feature points from visual data. This intermediary step filters out unreliable feature points caused by fast motion, thereby improving localization accuracy without significantly increasing overall system complexity
2Reliability
If base stations are erected in advance, then localization infrastructure is established, but deployment complexity increases
Solution Approach 1:
The patent extracts and removes the requirement for pre-deployed base stations from the localization system. By using visual feature points from the environment itself as localization references, the system eliminates the need for specialized infrastructure deployment while maintaining reliable localization through natural environmental features
Solution Approach 2:
The localization system uses the environment's own visual features as localization references, making the system self-sufficient without external base station infrastructure. The system serves itself by extracting feature points from captured images and using them for localization, eliminating deployment complexity
3Speed
If the object moves fast, then dynamic scenario capability is achieved, but localization accuracy deteriorates
Solution Approach 1:
The patent implements feedback mechanisms where inertial measurement unit (IMU) data continuously monitors motion state and feeds back to the feature point selection process. This feedback loop allows the system to identify and select reliable feature points even during fast motion, maintaining localization accuracy through real-time adaptation
Solution Approach 2:
The patent makes the feature point selection process dynamic by using inertial information to adaptively screen feature points based on current motion state. The system dynamically adjusts which feature points are used for localization depending on the object's motion characteristics, enabling accurate localization during fast movement
Data Source
AI summary
The present disclosure provides a localization method as well as a helmet and a computer readable storage medium using the same. The method includes: extracting first feature points from a target image; obtaining inertial information of the carrier, and screening the first feature points based on the inertial information to obtain second feature points; triangulating the second feature points of the target image to generate corresponding initial three-dimensional map points, if the target image is a key frame image; performing a localization error loopback calibration on the initial three-dimensional map points according to at least a predetermined constraint condition to obtain target three-dimensional map points; and determining a positional point of the specific carrier according to the target three-dimensional map points. In this manner, the accuracy of the localization of a dynamic object such as a person when moving can be improved.