Sensor Fusion Motion Map for Autonomous Driving
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing camera and radar-based measurement systems in autonomous driving lack comparability and accuracy in motion estimation, particularly in radial and angular directions, due to different data representations and uncertainties in measurements.
Innovation Solution
An apparatus and method that fuse motion information from multiple sensors, such as Doppler radar and stereo cameras, into a common representation, like a motion map, using error models and joint probability calculations to reduce uncertainty and enhance accuracy by aligning and combining data from different sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If motion information from multiple sensors is kept in different data representations, then each sensor maintains its native measurement characteristics, but comparability and integration between sensors deteriorates
Solution Approach 1:
The patent applies universality by creating a common motion representation framework that can accommodate multiple sensor types (stereo cameras, Doppler radar, LiDAR) with different native data representations. The system transforms various sensor outputs into a unified motion description format, enabling different sensors to serve multiple functions while maintaining their individual measurement strengths.
Solution Approach 2:
The patent employs parameter changes by transforming motion information from different sensor coordinate systems and data formats into a common representation. This involves changing parameters such as coordinate frames, velocity components (radial vs. angular), and uncertainty metrics to enable direct comparison and fusion of measurements from heterogeneous sensors.
2Loss of information
If stereo cameras are used for scene flow estimation, then high resolution and color information are obtained, but radial direction measurement accuracy deteriorates
Solution Approach 1:
The patent merges the complementary strengths of different sensors by combining stereo camera data (which provides excellent angular and texture information) with Doppler radar or LiDAR data (which provide accurate radial velocity measurements). This combination allows the system to maintain high-resolution texture and color information from cameras while compensating for their radial measurement limitations using radar or LiDAR.
3Measurement precision
If Doppler radar is used for motion measurement, then radial velocity accuracy is improved, but angular direction measurement accuracy deteriorates
Solution Approach 1:
The patent merges the complementary strengths of different sensors by combining Doppler radar data (which provides excellent radial velocity measurements) with stereo camera data (which provides angular position and texture information). This fusion allows the system to maintain accurate radial velocity measurements from radar while supplementing them with angular and visual information from cameras.
4Quantity of substance
If motion information from different sensors is combined without common representation, then comprehensive motion data is collected, but uncertainty and integration difficulty increase
Solution Approach 1:
The patent applies parameter changes by establishing a common motion representation that standardizes how motion information is expressed across different sensors. This involves transforming various sensor outputs into unified parameters such as position, velocity vectors, and covariance matrices in a common coordinate system, thereby reducing uncertainty through proper mathematical fusion while maintaining comprehensive motion data from multiple sources.
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
The solution provides a unified, accurate motion map with reduced uncertainty in both radial and angular directions, improving the precision of motion estimation and enabling better decision-making in autonomous driving systems.
Implementation Method 1
Doppler radar is a sensor system that uses radio waves to determine the velocity, range and angle of objects
Data Source
AI summary
An apparatus comprising circuitry configured to transfer motion information obtained from a plurality of sensors of different or similar type to a common representation.