In-Cabin Driver Posture Detection Using Standardized Image Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing complexity of vehicle operating environments and systems has led to higher driver risk, necessitating methods and systems for effectively detecting vehicle occupant actions.
Innovation Solution
A device and method that utilize previously classified image data from a database and current image data from vehicle interior sensors to detect vehicle occupant actions by comparing the two datasets, allowing for the generation of data representative of vehicle in-cabin insurance risk evaluations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If vehicle systems and operating environments become more complex to provide more functions and features, then vehicle functionality and entertainment capabilities are improved, but driver risk increases
Solution Approach 1:
The system performs preliminary classification of image data into multiple categories (driver actions, passenger actions, animal actions, object actions) before detailed analysis. This pre-categorization allows the system to anticipate and prepare for various risk scenarios, enabling faster response to potential safety issues while maintaining comprehensive monitoring of complex vehicle environments.
Solution Approach 2:
The image data is segmented into distinct categories representing different types of occupants and actions. By dividing the monitoring scope into separate classification categories (driver, passenger, animal, object) and further into specific actions, the system can analyze each segment independently, improving detection accuracy without being overwhelmed by the complexity of the entire scene.
2Measurement precision
If image data from vehicle interior sensors is compared with previously classified image data to detect occupant actions, then detection accuracy is improved, but processing time and computational complexity increase
Solution Approach 1:
Image data is pre-classified into multiple categories during or after capture, creating an organized structure of classified image data stored in databases. This preliminary classification eliminates the need for real-time categorization during comparison operations, significantly reducing processing time when detecting current occupant actions by only comparing against relevant pre-classified data.
Solution Approach 2:
The system creates copies of classified image data in structured databases organized by category (driver actions, passenger actions, animal actions, object actions). These copied and organized data sets can be quickly retrieved and compared without accessing the entire original data set, improving detection speed while maintaining accuracy through systematic comparison against relevant historical patterns.
Data Source
AI summary
A system comprising one or more processors and one or more non-transitory computer-readable media storing computing instructions that, when executed on the one or more processors, cause the one or more processors to perform operations including: receiving sensor data detected by one or more image sensors in a vehicle, wherein the sensor data is representative of driver movements of a driver in the vehicle; categorizing the sensor data as driver postures representative of actions of the driver in the vehicle; rotating and scaling the driver postures to be standardized for different drivers and for different locations of the one or more image sensors within different vehicles; analyzing the sensor data to determine a reference position of the driver in the vehicle; and storing, in a database, the driver postures, as rotated and scaled, and the reference position of the driver in the vehicle. Other embodiments are disclosed.