Monitoring Apparatus for Detecting Predetermined Actions in Inspection Areas
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing monitoring techniques are inadequate for effectively monitoring inspection areas, such as customs, due to congestion and the large number of people carrying luggage, which can lead to difficulties in detecting suspicious activities or changes in behavior.
Innovation Solution
A monitoring apparatus and system that includes a detection unit to identify individuals taking specific actions, such as being near luggage, making calls, changing visual lines, staying in one area, or entering/exit predetermined areas, based on images acquired from the inspection area, and an output unit to provide detection information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional monitoring techniques are used in inspection areas, then the monitoring coverage is provided, but the detection precision of suspicious activities deteriorates due to congestion and large number of people
Solution Approach 1:
The patent segments the monitoring task by detecting specific predetermined actions (squatting, crouching, making calls, staying in one area, entering/exiting) rather than attempting to monitor all activities. This segmentation allows the system to focus on high-risk behaviors that indicate suspicious activity, thereby improving detection precision in congested inspection areas where comprehensive monitoring is difficult.
Solution Approach 2:
The patent applies local quality by detecting specific actions at specific locations (inspection area) rather than requiring uniform monitoring of all activities. The system identifies suspicious behaviors localized to particular actions and positions, enabling precise detection of threats without needing to monitor every person's every action throughout the entire area.
2Reliability
If comprehensive monitoring of all activities is attempted, then the detection coverage is improved, but the system complexity increases due to congestion and large number of people
Solution Approach 1:
The patent extracts only the essential monitoring function needed for security - detecting predetermined suspicious actions. By taking out the complex task of monitoring all activities and retaining only the critical detection of predetermined actions (squatting, crouching, making calls, staying in one area, entering/exiting), the system achieves sufficient monitoring coverage while significantly reducing complexity.
Solution Approach 2:
The patent applies partial action by monitoring only specific predetermined actions rather than all possible activities. This partial monitoring approach is sufficient for security purposes, as it focuses on behaviors that indicate suspicious activity, thereby reducing system complexity while maintaining effective monitoring coverage.
3Measurement precision
If stress index monitoring is used, then suspicious persons can be identified, but false negatives occur when stress index is below threshold
Solution Approach 1:
Instead of monitoring for high stress levels and assuming low stress means innocence, the patent inverts the approach by directly detecting predetermined suspicious actions regardless of stress level. This inversion eliminates false negatives by focusing on observable suspicious behaviors rather than relying on stress thresholds, thereby improving detection reliability.
Solution Approach 2:
The patent replaces the physiological stress index measurement system with a visual action detection system. By substituting the complex physiological monitoring approach with direct observation of predetermined actions (squatting, crouching, making calls, staying in one area, entering/exiting), the system achieves more reliable detection that is not affected by stress level thresholds.
Data Source
AI summary
A monitoring apparatus (102) includes a detection unit (103) that detects a person P taking a predetermined action, based on an image being acquired by imaging an inspection area R, and an output unit (104) that outputs detection information relating to the person P being detected. For example, the predetermined action includes at least one of a first action being decided in relation to an item, a second action being decided in relation to a visual line of a person, a third action being decided in relation to a movement (including staying) of a person, a fourth action being decided in relation to entry in/exit from a predetermined area, and the like.