User Pattern Segmentation for Tracking Re-establishment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current inventory management systems in materials handling facilities face challenges in accurately tracking and identifying users as they move through the facility, particularly when they are out of the camera's view or in areas without coverage, leading to lost tracking and inefficiencies in inventory management.
Innovation Solution
A multiple-camera system that uses overhead cameras to capture color and depth values, establishing user patterns and descriptors to track users, and re-establish tracking by comparing newly detected patterns with stored data, ensuring continuous monitoring and inventory management even when users move out of view.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a single camera is used to track users, then the system is simple, but tracking is lost when users move out of camera view
Solution Approach 1:
The facility is divided into multiple zones, each monitored by overhead cameras positioned at different locations. Users are tracked by segmenting their movement path through these zones, with each camera capturing a portion of the overall tracking task. This resolves the contradiction by distributing the tracking function across multiple simple camera units rather than requiring one complex omnidirectional camera system.
Solution Approach 2:
A computing system acts as an intermediary that receives image data from multiple overhead cameras, processes the user patterns, and maintains continuous tracking across camera boundaries. This mediator coordinates the distributed camera system, enabling reliable continuous tracking while keeping individual camera units simple and overhead-mounted.
2Reliability
If overhead cameras are used to capture user patterns, then continuous monitoring is enabled, but data processing complexity increases
Solution Approach 1:
The system extracts only the essential user pattern features (color values, depth values, spatial coordinates) from the full image data captured by overhead cameras. By taking out only the necessary tracking-relevant information rather than processing complete images, the system achieves reliable user identification while reducing processing complexity.
Solution Approach 2:
User patterns are pre-processed and stored in a database before tracking operations. The system performs preliminary actions of capturing, processing, and storing user pattern data from overhead cameras during initial scans, so that during active tracking, the system can quickly retrieve and compare patterns without performing complex real-time processing, thus improving identification accuracy while managing complexity.
3Loss of time
If user patterns are stored in a database for comparison, then tracking can be re-established, but storage and processing requirements increase
Solution Approach 1:
The system stores user pattern data with differentiated quality and detail levels in the database. Frequently tracked users or users in critical zones have more detailed patterns stored, while less critical users have summarized patterns. This local quality differentiation enables faster tracking re-establishment for important users while reducing overall storage requirements by optimizing data retention based on local tracking needs.
Data Source
AI summary
Described is a multiple-camera system and process for re-identifying a user located in a materials handling facility based on user patterns and/or descriptors representative of the user. In one implementation, a user pattern and/or a plurality of descriptors representative of a user are maintained as a position of a user is tracked through a materials handling facility. If the tracking of the user is lost, the last known position is stored with the user pattern and/or descriptors. If a new object is detected and confirmed to be a user, a user pattern and/or descriptors of the new object are compared with the stored user pattern and/or descriptors to determine if the new object is the user.