Pedestrian Dead-Reckoning via Motion Pattern Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating pedestrian position indoors face challenges due to weak GPS signals and noise from mobile device sensors, making it difficult to provide accurate position information in environments where GPS signals are weak or unavailable.
Innovation Solution
A pedestrian dead-reckoning (PDR) technique based on pattern recognition using sensors in a user terminal, which recognizes specific pedestrian motions and applies unique PDR algorithms for accurate step, stride, and direction estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If GPS signals are used for position estimation, then position information can be obtained in outdoor environments, but the signal strength is weak and errors occur in indoor environments or areas with limited satellite visibility
Solution Approach 1:
The patent segments the position estimation problem into two distinct parts: GPS-based estimation for outdoor environments and sensor-based PDR estimation for indoor environments. The system selectively applies different estimation methods depending on the environment, thereby maintaining reliability across both settings while avoiding the precision limitations of GPS in indoor areas.
Solution Approach 2:
The patent introduces an intermediary system that uses mobile device sensors (accelerometer, gyroscope, magnetometer) as a mediator between the absence of GPS signals and the need for position estimation. This intermediary sensor-based PDR system bridges the gap in environments where GPS cannot provide reliable position information.
2Reliability
If mobile device sensors are used for pedestrian position estimation, then position information can be obtained indoors without GPS, but noise from device motion reduces measurement precision
Solution Approach 1:
The patent applies dynamic adjustment to the PDR algorithm by continuously updating step length and direction based on real-time sensor data and detected motion patterns. The system adapts to varying pedestrian behaviors and device orientations, dynamically adjusting parameters to maintain precision despite the noisy sensor environment.
Solution Approach 2:
The system incorporates feedback mechanisms where the detected motion patterns and calculated position estimates are continuously refined. The PDR algorithm uses feedback from sensor measurements to correct and improve position estimates, reducing the impact of noise through iterative refinement.
3Device complexity
If a single PDR algorithm is used for all pedestrian motions, then the system complexity is reduced, but the position estimation accuracy decreases for specific motion patterns
Solution Approach 1:
The patent segments the PDR algorithm into multiple specialized algorithms, each optimized for specific motion patterns such as walking, running, or using public transportation. The system detects the current motion pattern and selects the appropriate algorithm, thereby maintaining low overall complexity while achieving high precision for each specific motion type.
Solution Approach 2:
The system changes parameters based on detected motion patterns, adjusting step length, stride frequency, and direction calculation methods according to the specific activity. This parameter adaptation allows the system to maintain simplicity while achieving accurate position estimation across diverse pedestrian behaviors.
Data Source
AI summary
Disclosed are an apparatus for estimating a pedestrian position based on pedestrian motion recognition, and a method therefor. The method for estimating the pedestrian position based on pedestrian motion recognition includes recognizing a specific motion of a plurality of motions of the pedestrian, performing a unique pedestrian dead-reckoning (PDR) technique corresponding to the recognized specific motion among unique PDR techniques for each of the plurality of motions of the pedestrian, and estimating the pedestrian's position by the performed unique PDR technique.