Phase-Only Correlation for Vehicle Posture Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current self-driving systems face challenges in accurately determining vehicle posture changes due to uneven roads and acceleration, with existing solutions either being costly at the hardware level or time-consuming at the software level, particularly when using high-precision inertial navigation sensors or feature point matching methods.
Innovation Solution
The method employs a phase-only correlation technique to determine posture changes between road images captured by a traveling device, reducing costs and computational complexity while providing accurate posture information, and is applicable to monocular vision sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-precision inertial navigation positioning sensors are used to provide accurate posture information, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent replaces the mechanical/inertial sensing system with an optical image processing system. By using phase-only correlation methods on images captured by standard cameras, the system determines vehicle posture changes without requiring high-precision inertial navigation sensors, thus substituting a complex mechanical system with a simpler optical processing system.
2Measurement precision
If feature point matching is used to calculate posture changing, then measurement precision is improved, but productivity decreases due to time consumption
Solution Approach 1:
The patent changes the processing parameter from feature point coordinates to phase spectrum information. By transforming the image to the frequency domain and using phase-only correlation, the system directly calculates posture changes from phase differences, avoiding the time-consuming feature point matching process while maintaining accuracy.
Solution Approach 2:
The patent replaces the computational geometry approach of feature point matching with a frequency domain signal processing approach. The phase-only correlation method in the frequency domain provides a more efficient computational path for determining posture changes, significantly reducing calculation time.
3Productivity
If phase-only correlation method is used to determine posture changing, then productivity is improved by reducing computational complexity, but measurement precision may be affected
Solution Approach 1:
The patent extracts only the phase information from the frequency domain representation of the image, discarding the magnitude information. This phase-only extraction provides a simplified yet accurate measure of displacement and posture change, achieving high computational efficiency while maintaining measurement precision.
Data Source
AI summary
An image processing method and apparatus, and an electronic device, a storage medium, a computer program and a computer program product are provided. The method includes: at least two road image frames are obtained by an image collection assembly, which is arranged on a driving device; by using a phase correlation method, posture change information of the driving device between two road image frames among the at least two road image frames is determined; and posture information of the driving device is determined on based on the posture change information and reference posture information of the driving device.