Vehicle Occupant Safety Detection Using Mobile Camera Occlusion Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current seatbelt detection systems in vehicles are unable to accurately determine if the seatbelt is properly fastened and positioned across the occupant's torso due to occlusion by objects such as laptops or metal items, which can be missed by vehicle sensors like cameras and radar.
Innovation Solution
A system that utilizes vehicle sensors to capture first sensor data and, when occlusion occurs, requests second sensor data from mobile devices within the vehicle to ensure complete visibility of the occupant, employing image processing and artificial intelligence to determine seatbelt positioning and safety.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If vehicle sensors (cameras, radar) are used to detect seatbelt fastening status, then the detection system is simple and cost-effective, but the detection accuracy deteriorates when objects occlude the occupant's torso
Solution Approach 1:
The system leverages the mobile device's camera, which is already present for other purposes (entertainment, communication), and repurposes it for seatbelt detection. This multi-functional use of the mobile device camera avoids adding dedicated detection hardware while improving detection accuracy in occluded scenarios.
Solution Approach 2:
The mobile device acts as an intermediary between the vehicle's detection system and the occluded occupant. When the vehicle sensor cannot detect the seatbelt due to occlusion, the mobile device camera captures the obscured area, and the processor fuses this intermediate data with the original sensor data to determine seatbelt status.
2Measurement precision
If mobile device camera data is requested to overcome occlusion, then the detection accuracy improves, but the system complexity and communication requirements increase
Solution Approach 1:
The mobile device uses its own built-in camera to provide the additional detection data needed. The device serves itself by utilizing its existing imaging capability for the vehicle's safety detection purpose, eliminating the need for external or dedicated additional sensors.
Solution Approach 2:
The system merges data from two sources: the vehicle's original sensor and the mobile device's camera. The processor combines these data streams to form a complete view of the occupant and seatbelt, achieving higher detection accuracy through data fusion while maintaining relatively simple system architecture.
3Loss of information
If multiple sensor data sources are fused, then the completeness of occupant visibility improves, but the processing time and computational requirements increase
Solution Approach 1:
The system continuously monitors for occlusion conditions and pre-establishes communication protocols with mobile devices. When occlusion is detected, the request for additional sensor data is already prepared, reducing the time needed to acquire and process the supplementary information from the mobile device camera.
Solution Approach 2:
The detection process is segmented into distinct stages: initial detection by vehicle sensor, occlusion detection, conditional request for mobile device data, and data fusion. This segmentation allows the system to only perform additional processing when necessary, minimizing overall processing time while ensuring complete occupant visibility when needed.
Data Source
AI summary
Systems, methods, and other embodiments described herein relate to determining a safety issue relating to an occupant of a vehicle. In one embodiment, a method includes requesting, in response to a portion of the occupant being occluded in a first sensor data captured from a vehicle sensor in a vehicle cabin, a second sensor data from a mobile device capable of capturing second sensor data in the vehicle cabin. The method further includes determining, in response to the portion of the occupant that is occluded in the first sensor data being visible in the second sensor data, whether there is a safety issue.