Left Object Detection via Owner Identity Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing left object detection systems fail to accurately detect objects left behind in congested monitoring areas, particularly when the owner of the object moves out while another person remains nearby, leading to potential overlooking of the object.
Innovation Solution
A left object detection device that acquires images from a monitoring area, uses a processing section to identify static objects and track their ownership by comparing features with past images, and outputs a detection result after a predetermined time, ensuring accurate detection even in congested conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the monitoring area is congested with multiple objects and persons, then the complexity of image processing increases, but the detection accuracy of left objects deteriorates
Solution Approach 1:
The patent segments the monitoring process into distinct phases: object detection, owner identification, and left object determination. By dividing the complex monitoring task into manageable segments with specific criteria for each, the system handles congested areas more effectively without overwhelming processing complexity.
Solution Approach 2:
The system performs preliminary actions by detecting objects and identifying their owners before making the final determination of whether an object is left behind. This preliminary processing establishes ownership relationships in advance, simplifying the final detection decision and improving accuracy in congested environments.
2Reliability
If the system monitors multiple objects simultaneously in a congested area, then the monitoring coverage increases, but the non-detection error rate increases
Solution Approach 1:
The system uses feedback mechanisms by continuously monitoring whether persons remain near detected objects and updating the left object status accordingly. The timer mechanism provides feedback about the duration an object has been stationary, and the system adjusts its determination based on whether the original owner returns, reducing non-detection errors.
Solution Approach 2:
The system establishes ownership relationships as a preliminary action before final detection. By identifying which person owns which object in advance through image analysis and tracking, the system can accurately determine when an object is truly left behind, even when multiple objects are monitored simultaneously.
3Measurement precision
If the timer stops when any person is detected near the object, then false positives are reduced, but left objects are overlooked when the owner remains nearby
Solution Approach 1:
The system applies local quality by differentiating between different types of persons near an object. Instead of treating all persons equally, it specifically identifies whether the person is the owner (based on preliminary identification) or another person, applying different timer stopping rules accordingly. This localized differentiation improves both precision and reliability.
Solution Approach 2:
The system changes the parameter of timer behavior based on the identity of the person detected. When the owner is detected, the timer continues running; when another person is detected, the timer stops. This parameter change based on person identity allows the system to distinguish between temporary absence and genuine left objects.
Data Source
AI summary
A left object detection device for detecting an object left behind in a monitoring target area includes an acquisition section which acquires an image of the area from an image pickup device for picking up an image of the monitoring target area, a processing section, and an output section which outputs in accordance with a result of the processing section. Upon detection of the object which has been kept static for a given time period as an article from the image acquired by the acquisition section, the processing section starts a timer for counting time from detection of the article, determines whether or not a person who owned the article matches a person in the image acquired by the acquisition section, stops the timer upon determination that they match, and does not stop the timer upon determination that they do not match. The processing section determines whether or not the timer counts in excess of a predetermined time period. Upon determination made by the processing section that the timer counts in excess of the predetermined time period, the output section outputs an existence of the left object.