Pedestrian Gait Recognition for Continuous Robotic Assistance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing assistive robotic devices struggle to maintain continuity of assistance for mobile pedestrians due to the inability to distinguish between individuals without wearable identification, leading to inefficient reinteraction and compromised service delivery, especially in public spaces.
Innovation Solution
Utilizing pedestrian gait recognition through visual data analysis to identify and track pedestrians across different locations, enabling continuous assistance based on their unique needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If facial recognition is employed to identify mobile pedestrians, then pedestrian identification accuracy is improved, but user privacy intrusion increases and system complexity increases
Solution Approach 1:
The patent replaces facial recognition (optical/biological system) with gait recognition (mechanical motion analysis). By analyzing the mechanical patterns of walking - stride length, cadence, arm swing, and body posture - the system achieves pedestrian identification without capturing facial features, thus reducing privacy intrusion while maintaining identification capability
Solution Approach 2:
The patent extracts and analyzes only the essential motion characteristics from visual data - specifically gait parameters such as stride length, walking speed, and body posture - while discarding other unnecessary visual information. This extraction approach simplifies the processing requirements and reduces system complexity while maintaining identification accuracy
2Reliability
If each mobile pedestrian must repeat their previous interaction with subsequently encountered assistive robotic devices, then identification reliability is improved, but time efficiency deteriorates
Solution Approach 1:
The patent performs preliminary gait pattern analysis and stores the pedestrian's unique gait signature during the first interaction. When the pedestrian encounters another robotic device, the system retrieves and compares the stored gait pattern with the observed gait, enabling rapid reidentification without requiring the pedestrian to repeat their interaction or provide information again
Solution Approach 2:
The patent creates a digital copy of the pedestrian's gait pattern characteristics - capturing the unique mechanical signature of their walking style - and stores this copy for future recognition. This gait template serves as a reusable identifier that can be quickly compared against new visual data to confirm pedestrian identity across multiple encounters
3Measurement precision
If wearable identification devices are used to distinguish mobile pedestrians, then pedestrian identification accuracy is improved, but device complexity and user burden increase
Solution Approach 1:
The patent enables the pedestrian's own body to serve as the identification carrier by analyzing the inherent mechanical characteristics of their gait. The robotic devices capture visual data of the pedestrian walking and automatically extract gait parameters such as stride length, cadence, and posture variations, eliminating the need for any wearable identification devices while maintaining high identification accuracy
Solution Approach 2:
The patent shifts the identification basis from static parameters (facial features requiring close proximity) to dynamic parameters (gait characteristics that can be measured from a distance). By analyzing temporal and spatial variations in walking patterns - such as stride frequency, step length variability, and body sway - the system achieves reliable identification without requiring pedestrians to wear any additional devices
Data Source
AI summary
A a method for robotic assistance continuity to a pedestrian using gait recognition. The method includes determining a gait pattern of the pedestrian based on analysis of visual data obtained at a first position. The pedestrian is provided assistance based on a query at the first position. The pedestrian is identified at a subsequent position by identifying the gait pattern from analysis of visual data obtained at the subsequent position. The need for additional pedestrian assistance is determined at the subsequent position.