In-Vehicle Event Recording for Driver Inattention Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in accurately determining the causal relationship between driver inattention and vehicle accidents, particularly when occupants' face orientations are similar, making it difficult to identify the cause of accidents.
Innovation Solution
An on-vehicle recording control apparatus and method that includes video data acquisition, orientation detection, and event detection units to capture and analyze video data from both external and internal vehicle cameras, detecting when the driver's face orientation deviates from the vehicle's travel direction and correlating this with vehicle acceleration events to initiate event recording.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the system records video data only when face orientations of occupants are the same, then the recording trigger is simplified, but the accuracy in determining causal relationship between driver inattention and accidents is reduced
Solution Approach 1:
The system segments the monitoring parameters into multiple independent dimensions: face orientation detection, line of sight detection, acceleration detection, and inter-vehicle distance detection. By dividing the complex causal determination into separate detectable parameters, the system achieves both operational simplicity and determination accuracy.
Solution Approach 2:
The system transitions from two-dimensional face orientation matching to multi-dimensional monitoring by adding line of sight direction, vehicle acceleration, and inter-vehicle distance as additional detection dimensions. This enables comprehensive causal analysis through correlated parameters across multiple dimensions.
2Measurement precision
If the system monitors multiple parameters (face orientation, line of sight, acceleration, inter-vehicle distance), then the causal relationship determination is improved, but the system complexity increases
Solution Approach 1:
The control device integrates multiple detection functions into a single universal system that simultaneously monitors face orientation, line of sight, acceleration, and inter-vehicle distance. This multi-functional integration reduces overall system complexity compared to separate independent monitoring systems.
Solution Approach 2:
The system merges the detection of driver state (face orientation and line of sight) with vehicle state (acceleration and inter-vehicle distance) into a unified monitoring framework. By combining these detection functions, the system achieves comprehensive causal analysis without proportionally increasing complexity.
3Loss of information
If the system stores video data from event recording start flag to event detection time point, then the complete accident context is captured, but the data storage requirement increases
Solution Approach 1:
The system performs preliminary actions by detecting and flagging abnormal states (face orientation deviation, line of sight deviation, abnormal acceleration, short inter-vehicle distance) before accidents occur. Video data is stored from these preliminary abnormal state detections to the accident event time point, capturing complete context while limiting storage to periods following abnormal detections rather than continuous storage.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
To appropriately record an event, such as an accident, which is caused by inattention of a driver of a vehicle, in addition to recording a cause of the event. A control device 100 includes a video data acquisition unit 120, an orientation detection unit 131 that determines whether a first condition that the driver faces a direction other than a traveling direction of the vehicle is met, an event detection unit 132, and a recording control unit 133 that, if an event is detected, stores first video data including at least an event detection time point as event recording data. The control device 100 performs a process a), a process b), or a combination of the process a) and the process b) when the first condition is met, where a) the recording control unit 133 adds an event recording start flag to the first video data, and if an event is detected while the first condition is met, the recording control unit 133 stores the first video data from the event recording start flag to at least the event detection time point as the event recording data, and b) the event detection unit 132 reduces a threshold for the acceleration for detecting an event and then detects an event.