Visual Pose Determination in GPS-Denied Navigation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Traditional GPS systems and triangulation methods fail in GPS-denied environments, such as indoors or areas with sky blockage, leading to inaccurate location determination and inefficient navigation, especially in vehicle transportation management and car-for-hire services.

Innovation Solution

A visual mapping and transportation management system using a mobile computing device's camera to capture images, which are then processed by an image-matching server to determine the device's location and direct vehicles, leveraging a map framework of linked keyframes and interest features for accurate positioning without relying on GPS or power-intensive triangulation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If GPS receivers are used to determine location, then location information can be obtained, but accuracy deteriorates in GPS-denied environments such as indoors or areas with sky blockage

Engineering Contradiction:
Improvelocation determination reliabilityVSAvoidlocation accuracy
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The patent introduces visual features and image matching as an intermediary mechanism to bridge the gap when GPS fails. Instead of directly relying on satellite signals, the system uses camera-captured images, extracts visual features, and matches them against a pre-built map framework to indirectly determine location, thereby maintaining reliability in GPS-denied environments

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces the GPS satellite-based electromagnetic signal system with a visual-based location determination system. By substituting the mechanical/electromagnetic GPS reception mechanism with optical image capture and processing, the system achieves location accuracy in environments where traditional GPS mechanics fail due to signal blockage

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If triangulation methods are used to determine geolocation, then location can be obtained, but the process becomes power-intensive and less efficient

Engineering Contradiction:
Improvelocation determination capabilityVSAvoidbattery consumption
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent applies preliminary action by pre-building and storing a comprehensive map framework with visual features before the actual navigation task. This preprocessing work is done when energy availability is not constrained, and then the stored visual map is used for efficient matching during navigation, avoiding the need for real-time triangulation calculations that consume excessive power

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent creates a visual copy of the physical environment through pre-captured images and extracted features stored in a map framework. Instead of performing computationally intensive real-time triangulation, the system compares and matches this visual copy against current camera images, significantly reducing processing power and battery consumption while maintaining location determination capability

Inventive Principle:
Principle #26Copying

3Reliability

If GPS and triangulation systems are used for navigation, then location information is provided, but user interaction increases and navigation becomes less intuitive

Engineering Contradiction:
Improvenavigation functionalityVSAvoiduser interaction requirement
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The patent implements self-service by enabling the navigation system to automatically capture images, extract visual features, match them against the stored map, and determine location without requiring user intervention. The system autonomously performs the entire visual-based location determination process, eliminating the need for users to manually input data or configure settings, thereby improving ease of operation while maintaining navigation reliability

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS9964955B2System and methods of determining a geometric pose of a camera based on spatial and visual mapping
Publication Date: 2018.05.08 AMAZON TECH INC

AI summary

The disclosure provides systems and methods of use pertaining to a visual mapping and transportation management system for determining a location of a user and directing a vehicle to the user's location. Embodiments include a navigation application installed upon a user's mobile computing device and configured to transmit a user image from the device to an image-matching server storing a map composed of keyframes, each having a stored image, a known geometric pose, and numerous extracted interest features. The server also includes a processor configured to extract interest features from the user image, compare the interest features between the user image and the stored images, identify common interest features between the two, and based on the common interest features and known geometric poses of the stored images, determine a global geometric pose of the user image before directing the vehicle to the user's location. Other embodiments are also disclosed.