Multi-Camera Person Identification Using Position Correspondence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing monitoring camera systems struggle to identify objects without markers in overlapping areas, leading to difficulties in re-identification tasks and increased processing load and accuracy issues due to coordinate conversion errors.
Innovation Solution
The system generates a correspondence relationship between image elements of multiple cameras based on detected object positions, using feature information and positional correspondence without markers, enabling robust re-identification through collation of feature information and positional relationships.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If coordinate conversion parameters are used to convert object positions to a common coordinate system, then objects in overlapping areas can be identified, but the system becomes difficult to operate without markers and processing load increases
Solution Approach 1:
The patent extracts and utilizes the detected object positions themselves as the basis for establishing correspondence relationships between cameras, eliminating the need for external markers. By taking the object's position in each camera's coordinate system as the key feature, the system achieves markerless operation while maintaining identification accuracy.
Solution Approach 2:
The patent introduces a correspondence relationship table as an intermediary structure that maps object positions between different camera coordinate systems. This intermediary mechanism enables the system to convert and compare positions across cameras without requiring physical markers or complex coordinate conversion parameters.
2Reliability
If coordinate conversion is performed to identify objects in overlapping areas, then re-identification is enabled, but processing load and coordinate conversion errors increase
Solution Approach 1:
The patent creates a simplified copy of the position correspondence relationship by directly storing mapped positions in a correspondence table rather than performing complex real-time coordinate conversions. This copying approach reduces processing load while maintaining re-identification reliability.
Solution Approach 2:
The patent performs preliminary action by pre-establishing correspondence relationships between camera coordinates and storing them in advance. This allows the system to avoid heavy computational coordinate conversion during actual re-identification tasks, reducing processing load while maintaining accuracy.
3Manufacturing precision
If markers are used to establish coordinate relationships, then position conversion is accurate, but the system complexity and installation requirements increase
Solution Approach 1:
The patent enables the system to be self-service by automatically detecting object positions and using them to establish correspondence relationships between cameras. This eliminates the need for external markers and manual installation, reducing device complexity while maintaining position conversion accuracy through the self-generated correspondence table.
Data Source
AI summary
An identification method includes generating, when a first person is detected from a first image captured by a first camera and a second person is detected from a second image captured by a second camera, relationship information obtained by associating a position in the first image from which the first person is detected with a position in the second image from which the second person is detected, and first identifying, based on feature information on the first person and feature information on the second person, the first person and the second person, by a processor.