Surveillance Group Detection Using Temporal Tracking to Filter False Alerts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing systems struggle to differentiate between normal and suspicious groups of people in surveillance footage, leading to an excessive number of alerts and increased complexity in surveillance, as they often detect gatherings that are not inherently suspicious.
Innovation Solution
An image processing system that detects groups in surveillance footage, determines if the group has been previously identified, and provides an alert only when the group exhibits a high degree of abnormality by analyzing features such as similarity in appearance and frequency of occurrence, thereby reducing false alarms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the system detects all groups in surveillance footage, then the detection coverage is improved, but the number of false alerts increases and surveillance complexity increases
Solution Approach 1:
The system performs preliminary actions by detecting and tracking groups across multiple time points before making a final determination. The determination unit checks whether a detected group has been detected in the past, requiring temporal verification before triggering an alert. This preliminary tracking and comparison process filters out transient or normal gatherings, reducing false alerts while maintaining comprehensive detection coverage.
2Speed
If the system provides alerts for all detected groups, then the alert responsiveness is improved, but the number of false alerts increases
Solution Approach 1:
The system implements feedback by continuously monitoring whether detected groups have been previously identified. The determination unit uses historical detection data to verify the abnormality of current groups, creating a feedback loop that distinguishes between normal and suspicious gatherings. This feedback mechanism maintains rapid alert responsiveness for genuine threats while filtering out false positives through comparative analysis.
Data Source
AI summary
To provide an image processing system, an image processing method, and a program, capable of detecting a group with high irregularity. An image processing system is provided with: a group detector that detects a group based on an input image captured with an image capturing at a first time; a repeating group analyzer that determines that a detected group has been previously detected; and an alert module that reports when the detected group has been determined by the repeating group analyzer to have been previously detected.