3D Gaze Vector Calculation for Driver Attention Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current gaze detection systems in vehicles can only determine the direction of a driver's gaze but not what captures their attention in a three-dimensional space, failing to assess attention to specific points or objects outside the vehicle effectively.
Innovation Solution
A system that calculates gaze vectors in a three-dimensional space based on vehicle location, orientation, and gaze direction, determining driver attention to points or objects and using this information to direct the driver's attention through notification systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If general gaze direction detection is used, then the system can determine where the driver is looking, but it cannot determine what the driver is actually attending to in three-dimensional space
Solution Approach 1:
The patent transforms two-dimensional gaze detection into three-dimensional attention detection by incorporating vehicle location and orientation data. The system calculates gaze vectors in 3D space by combining camera-derived gaze directions with vehicle pose information, enabling determination of what objects the driver is attending to rather than just where they are looking on a screen or dashboard.
Solution Approach 2:
The patent introduces vehicle location and orientation as intermediary parameters that bridge the gap between gaze direction and attention object. By using the vehicle's pose as a mediator, the system can transform abstract gaze vectors into concrete spatial coordinates that identify specific objects in the environment, such as road signs or pedestrians.
2Device complexity
If the system tracks only general gaze direction, then the implementation is simpler, but it fails to provide accurate attention assessment for specific points or objects outside the vehicle
Solution Approach 1:
The patent makes the existing gaze detection system multi-functional by adding attention determination capabilities. The same camera and processing units that detect gaze direction are extended to also calculate three-dimensional gaze vectors and determine attention to external objects, eliminating the need for separate hardware while enhancing functionality.
Solution Approach 2:
The patent changes the parameter representation from simple gaze direction angles to three-dimensional gaze vectors that incorporate vehicle location and orientation. This parameter transformation enables the system to reliably identify external objects while using the same computational resources, thereby improving reliability without proportionally increasing complexity.
3Productivity
If the system makes general determinations of driver gaze location, then the processing is faster, but it cannot provide detailed information about driver attention to specific objects for safety applications
Solution Approach 1:
The patent performs preliminary calculations by pre-establishing the relationship between vehicle pose and gaze vector transformation. The system prepares the coordinate transformation matrices and spatial reference frames in advance, allowing rapid conversion of gaze directions to three-dimensional attention coordinates without real-time computational overhead, thus maintaining processing speed while enabling detailed object identification.
Data Source
AI summary
Methods and systems are provided for detecting an attention of an occupant of a vehicle. In one embodiment, a method includes calculating, by a processor, a first gaze vector in a three-dimensional space based on a first vehicle location, a first vehicle orientation, and a first gaze direction; calculating, by the processor, a second gaze vector in the three-dimensional space based on a second vehicle location, a second vehicle orientation, and a second gaze direction; and determining the attention of the occupant based on the first gaze vector and the second gaze vector.