Polarimetric Road Image Normalization for Free Space Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computer vision systems for autonomous vehicles rely on color video data, which is suboptimal for accurately estimating free space due to similarities in color between drivable and non-drivable surfaces, leading to reduced accuracy in path planning.
Innovation Solution
The use of polarimetric image data normalized by calculating the Sun location using a GPS receiver and compass, allowing the electronic control unit (ECU) to normalize the polarimetric images such that the angle of linear polarization (AoLP) and degree of linear polarization (DoLP) remain at predetermined fixed values, effectively isolating the road surface in the imaged drive scene.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If color video data is used for free space estimation, then hardware costs are reduced, but measurement precision deteriorates due to color similarities between drivable and non-drivable surfaces
Solution Approach 1:
The patent transitions from color-based (2D RGB) image data to polarimetric (3D polarization state) image data by capturing the angle of linear polarization and degree of linear polarization dimensions. This additional dimensional information enables differentiation of surfaces with similar colors but different polarization properties, resolving the measurement precision issue without requiring complex multi-sensor systems.
Solution Approach 2:
The patent changes the fundamental parameter used for surface differentiation from color (wavelength) to polarization state (angle and degree of linear polarization). By measuring how light polarizes upon reflection from different surfaces, the system achieves superior free space estimation accuracy compared to color-based methods, while maintaining a single-camera architecture.
2Reliability
If polarimetric image data is used without normalization, then free space estimation accuracy is improved, but reliability deteriorates due to varying Sun locations affecting polarization measurements
Solution Approach 1:
The patent performs preliminary normalization of polarimetric image data by calculating the Sun's position using GPS and compass data, then adjusting the polarization angle representation accordingly. This preliminary action compensates for varying illumination conditions before the data is used for free space estimation, ensuring consistent and reliable results across different times of day and weather conditions.
Solution Approach 2:
The system uses GPS receiver and compass data as feedback inputs to dynamically adjust the polarimetric image normalization process. By continuously monitoring the vehicle's position and orientation, the system calculates the Sun's relative position and applies appropriate normalization transformations, creating a closed-loop system that maintains reliability under varying environmental conditions.
3Measurement precision
If GPS receiver and compass are added to calculate Sun location, then polarimetric image normalization is improved, but device complexity increases
Solution Approach 1:
The patent leverages the GPS receiver and compass, which are standard navigation components in modern vehicles, to serve the additional function of Sun position calculation for polarimetric normalization. By reusing existing multi-functional components rather than adding dedicated Sun-tracking hardware, the system achieves accurate Sun location estimation without proportionally increasing device complexity.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach improves the accuracy of free space estimation and path planning by reducing the impact of changing Sun locations on polarimetric representations, enhancing the overall drive experience and reducing hardware costs associated with perception tasks.
Implementation Method 1
a polarimetric camera to collect polarimetric image data of a drive scene of the host vehicle, with the drive scene including the road surface and its surrounding environs
Data Source
AI summary
A system for a host vehicle operating on a road surface includes a polarimetric camera, a global positioning system (“GPS”) receiver, a compass, and an electronic control unit (“ECU”). The camera collects polarimetric image data of a drive scene, including a potential driving path on the road surface. The ECU receives the polarimetric image data, estimates the Sun location using the GPS receiver and compass, and computes an ideal representation of the road surface using the Sun location. The ECU normalizes the polarimetric image data such that the road surface has a normalized representation in the drive scene, i.e., an angle of linear polarization (“AoLP”) and degree of linear polarization (“DoLP”) equal predetermined fixed values. The ECU executes a control action using the normalized representation.