Unattended Object Monitoring With Mask-Area Alarm Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing surveillance systems issue alarms for unnecessary unattended objects, increasing the burden on observers, particularly in large areas or when there are many such objects.
Innovation Solution
An unattended object monitoring device that tracks objects in a monitoring area, detects unattended objects that do not move for a predetermined time, and excludes them from alarms if they are within a preset mask area or of a specific type associated with the mask area.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the surveillance system issues alarms for all detected unattended objects, then the detection coverage is improved, but the observer's burden increases and false alarms increase
Solution Approach 1:
The monitoring area is segmented into multiple mask areas, each corresponding to a specific type of unattended object (e.g., store signs, construction cones, information boards). By dividing the space into distinct zones, the system can selectively apply exclusion rules to specific areas, allowing comprehensive monitoring while filtering out known benign objects in designated zones.
Solution Approach 2:
Mask areas are preset in advance based on the types of unattended objects that are expected to appear in the monitoring area. This preliminary configuration allows the system to proactively identify and exclude certain objects before they trigger false alarms, reducing the observer's burden while maintaining detection coverage for suspicious objects.
2Area of stationary object
If the monitoring area is expanded to cover larger spaces, then the surveillance coverage is improved, but the number of unnecessary alarms increases
Solution Approach 1:
The expanded monitoring area is divided into multiple mask areas, each tailored to specific object types. This segmentation allows the system to manage large spaces efficiently by applying localized exclusion rules only where appropriate, rather than uniformly across the entire area.
Solution Approach 2:
Different regions of the monitoring area are assigned different qualities or characteristics through mask area definitions. Certain zones are designated for specific object types, allowing the system to apply context-aware filtering that adapts to the local characteristics of each area, thereby reducing unnecessary alarms in regions where certain objects are expected.
3Measurement precision
If the system monitors all objects in the monitoring area, then the detection accuracy is improved, but the processing complexity increases
Solution Approach 1:
The system extracts and isolates the mask area information from the overall detection process. By separating the exclusion logic into distinct mask area definitions, the system can efficiently filter out known benign objects without complicating the core detection algorithm, thereby maintaining detection accuracy while reducing processing complexity.
Solution Approach 2:
Mask areas are predetermined and stored in advance, allowing the system to quickly reference pre-defined exclusion zones during real-time monitoring. This preliminary preparation eliminates the need for complex real-time analysis of object types in every detected instance, simplifying the processing while maintaining accurate detection and exclusion capabilities.
Data Source
AI summary
Provided is an unattended object monitoring device for detecting an unattended object in a monitoring area and issuing an alarm therefor, while excluding a detected object for which issuance of an alarm is unnecessary from alarm target objects for which alarms are issued. An unattended object monitoring device is configured to: detect an object appearing in a monitoring area from images of the area captured by an image pickup device to track each detected object between the captured images; detect, as an unattended object, an object that does not move for more than a predetermined time based on a tracking result; determine if the detected unattended object is present in a mask area preset in the monitoring area; and exclude the detected unattended object which is present in the mask area from alarm target objects for which alarms are issued.