Surface Image Registration for GPS-Free Object Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing localization methods for objects, such as vehicles or robots, face challenges in accurately determining the location of moving objects, especially in environments without standard positioning systems like GPS, and in real-time motion scenarios.
Innovation Solution
The method involves using an imaging sensor to capture sequential images of a surface that overlap with previous images, registering the current image with the previous one, and computing the current location of the object relative to a reference location based on the analysis of the registration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If standard localization approaches (landmark recognition or GPS) are used, then location determination is straightforward in familiar environments, but localization fails or becomes inaccurate in environments without known landmarks or GPS availability
Solution Approach 1:
The system creates its own reference framework by capturing images of the environment and generating synthetic views from those images. Instead of relying on external GPS or pre-existing landmark databases, the system serves itself by building localization references from the captured environmental data, enabling operation in GPS-denied and unfamiliar environments
Solution Approach 2:
The system creates synthetic images that copy or simulate views from different locations based on captured reference images. These synthetic images serve as virtual landmarks or reference points for localization, allowing the system to match current sensor data against synthesized references to determine position without requiring actual physical landmarks or GPS signals
2Productivity
If image registration and analysis are performed in real-time during motion, then continuous localization is achieved, but computational complexity and processing time increase
Solution Approach 1:
The system performs preliminary actions by capturing and storing reference images of the environment before localization is needed. These pre-captured images are used to generate synthetic reference views offline, reducing the computational burden during real-time operation to only matching current sensor data against pre-generated synthetic references rather than performing full image registration from scratch
Solution Approach 2:
The system introduces synthetic images as an intermediary between captured reference images and real-time localization matching. The synthetic images serve as a simplified intermediate representation that captures essential geometric and visual features, making the real-time matching process computationally more efficient while maintaining localization accuracy
Data Source
AI summary
There is provided a computer implemented method of controlling movement of an object, comprising: accessing a current image of a surface relative to an object at a current location, wherein an imaging sensor is set to capture the current image depicting an overlap with a previously captured image of the surface when the object was at a previous location, registering the current image to the overlap of the previously captured image, computing the current location of the object relative to a reference location according to an analysis of the registration, and feeding the current location into a controller for controlling movement of the object.