Feature Selection for Image-Based Indoor Location
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mobile devices face challenges in determining location in areas without GPS or WiFi coverage, such as indoor spaces or remote locations, as they lack compatible modules or have unavailable geographic location systems.
Innovation Solution
The method involves using a camera image to compare with pre-stored geometry data to identify features and determine location, ranking features by uniqueness, and generating geometry data for localization, which can be filtered based on user characteristics and device limitations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If GPS or WiFi-based location methods are used, then location determination is available, but the system becomes dependent on network coverage and satellite information that may be unavailable in certain locations
Solution Approach 1:
The patent introduces image data as an intermediary medium between the mobile device and location determination. Instead of directly using GPS or WiFi signals, the system captures images of the environment and compares them against pre-stored reference images and geometry data to infer location, thereby eliminating dependency on network coverage or satellite signals.
Solution Approach 2:
The patent replaces the traditional mechanical/electromagnetic location determination systems (GPS satellites, WiFi networks) with a visual recognition system that uses camera images and pattern matching. This substitution allows location determination through visual features rather than electromagnetic signals, enabling operation in environments where traditional systems fail.
2Measurement precision
If all identified features are used for location determination, then location accuracy is improved, but the amount of data to be processed and stored increases significantly
Solution Approach 1:
The patent applies local quality by differentiating features based on their uniqueness and relevance to location determination. Instead of treating all features equally, the system ranks features according to their distinctiveness and selects only those that provide meaningful location information, thereby reducing data volume while maintaining accuracy.
Solution Approach 2:
The patent extracts only the essential features from the complete set of identified features. By filtering and selecting features based on uniqueness criteria, the system removes redundant or less informative features, retaining only the necessary geometry data for accurate location determination without excessive data storage requirements.
3Measurement precision
If features are selected based on user characteristics, then location determination becomes more personalized and accurate, but the system complexity increases
Solution Approach 1:
The patent implements preliminary action by pre-defining user characteristics and their corresponding feature selection criteria before actual location determination occurs. The system includes pre-configured rules that map user types to specific feature priorities, eliminating the need for complex real-time analysis and reducing operational complexity while maintaining personalization.
Data Source
AI summary
Feature selection is provided for geometry data in an image-based based location determination. For example, one or more computing devices, may receive information collected at a particular area. The one or more computing devices may identify one or more features and associated locations from the received information. The identified one or more features may be ranked according to relative uniqueness among the identified one or more features. A set of geometry data for locating a mobile device at the particular area is generated based on selected ones of the ranked features and the associated locations.