Person Search System Using Pre-analyzed Visual Similarity Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current security systems are inefficient in identifying and tracking individuals visually similar to criminals at crowded locations, leading to delayed initial investigations due to the time-consuming process of collecting and analyzing camera footage and eyewitness reports.
Innovation Solution
A person search system that connects multiple cameras to a server and client terminal, allowing for the analysis and comparison of positional and characteristic information across multiple camera feeds to efficiently identify and display similar individuals, facilitating a prompt grasp of a criminal's visual characteristics and escaping direction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple camera feeds are manually analyzed to identify visually similar individuals, then identification accuracy is improved, but investigation time increases significantly
Solution Approach 1:
The system performs preliminary analysis by pre-processing video data from multiple cameras and pre-calculating visual similarity metrics before an actual investigation request is made. This allows the system to quickly retrieve pre-analyzed results during urgent investigations, resolving the contradiction between thorough analysis and time efficiency.
Solution Approach 2:
The system creates visual similarity models (digital copies) of individuals from camera footage and uses these models to efficiently search for matching individuals across multiple camera feeds. This copying approach enables rapid identification without manually reviewing all original video data, maintaining accuracy while reducing investigation time.
2Reliability
If comprehensive analysis of position information and characteristic information from all cameras is performed, then identification reliability is improved, but system complexity increases
Solution Approach 1:
The system segments the identification process into distinct modules: position information analysis, characteristic information analysis, and visual similarity comparison. Each module handles a specific aspect of the analysis independently, making the overall system more manageable and easier to implement while maintaining comprehensive analysis capabilities.
Solution Approach 2:
The system introduces an intermediary visual similarity model that mediates between raw camera data and final identification results. This intermediary representation simplifies the comparison process by reducing complex video data into key visual features, thereby reducing system complexity while preserving identification reliability.
3Loss of information
If visual similarity comparison is performed across all camera feeds simultaneously, then search completeness is improved, but processing speed decreases
Solution Approach 1:
The system dynamically adjusts the search process by first performing quick visual similarity comparisons across all cameras to identify potential matches, then focusing detailed analysis only on promising candidates. This dynamic approach ensures comprehensive search coverage while maintaining high processing speed by avoiding unnecessary detailed analysis of all camera feeds.
Data Source
AI summary
When captured videos transmitted from n units of cameras are received, a server analyzes position information and characteristic information on a person appearing in each of captured videos as analysis results. A client terminal displays, on a screen, each of a k number of captured videos that are selected by a user operation. According to a selection of a search target person, the client terminal transmits, to the server, a search request for the search target person in a (k−1) number of captured videos other than the captured video in which the search target person appears. The server extracts the search target person and transmits the analysis result to the client terminal. The client terminal displays cut-out images of one or more persons identical with or similar to the search target person based on the analysis result.