External Camera Occupant Alerts for Left-Behind Child Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle alert systems are inadequate in reliably detecting and alerting for left-behind children, particularly small children who may be in blind spots within the vehicle, as they rely on door opening/closing history or internal image comparison which can be inconveniences and inefficient.
Innovation Solution
An alert system utilizing external-environment cameras to capture images of the vehicle's exterior, combined with in-vehicle sensors, to determine the presence or absence of children by analyzing changes in images and outputting alerts when a child is likely left behind, ensuring accurate detection and minimizing false alarms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If door opening/closing history is recorded to detect left-behind occupants, then alert function is provided, but detection reliability is insufficient for small children in blind spots
Solution Approach 1:
The patent transitions from monitoring door events (1D temporal sequence) to using 3D spatial image data from external cameras. By capturing images from multiple external cameras and analyzing spatial relationships between detected objects and vehicle blind spots, the system achieves reliable detection of small children who would be missed by door-history-based methods alone.
2Measurement precision
If internal image comparison is used to detect occupants, then detection capability is provided, but false alerts increase and user convenience decreases
Solution Approach 1:
The patent introduces external environment cameras as intermediary detection devices that observe the vehicle exterior and detect objects in blind spots. This external perspective serves as a mediator between the occupant detection need and the internal camera system, providing complementary detection data that reduces false alerts by cross-validating with independent external observations.
Solution Approach 2:
The system merges detection results from multiple sources: door opening/closing history, internal environment camera images, and external environment camera images. By combining these diverse detection modalities and analyzing their consistency, the system achieves higher detection accuracy while reducing false alerts through multi-source verification.
3Reliability
If multiple detection methods are combined, then detection reliability improves, but system complexity increases
Solution Approach 1:
The external environment cameras serve multiple functions: they capture images for left-behind occupant detection, provide spatial information for blind spot analysis, and enable object recognition across different vehicle zones. This multi-functionality reduces the need for separate dedicated sensors for each detection task, thereby managing system complexity while maintaining high reliability.
Data Source
AI summary
An alert system configured to output an alert about an occupant being left in a vehicle includes an external-environment camera, a determiner, and an alert controller. The external-environment camera captures images at least one of which includes the occupant outside the vehicle. The determiner determines whether a child including an infant is likely to have been left in the vehicle. The alert controller outputs an alert if the child is likely to have been left in the vehicle. The determiner determines whether the child is likely to have been left in the vehicle, based on a change in the images from one of a presence state and an absence state of a person or an object outside the vehicle in a first image of the images to another of the presence state and the absence state of the person or the object in a second image of the images.