Four-Fold Symmetric Image Filtering for Accurate Stereo Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Stereo matching accuracy is reduced due to noise, particularly zipper noise, in images generated by stereo cameras used for vehicle driving assistance systems.
Innovation Solution
An image processing device employs a filter pattern with alternating positive and negative polarity coefficients, having four-fold rotational symmetry, to perform filtering on left and right images, reducing zipper noise while maintaining object features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If conventional filtering methods are used to reduce zipper noise, then noise reduction is achieved, but stereo matching accuracy deteriorates due to loss of object features
Solution Approach 1:
The filter pattern applies different coefficients to different regions of the image. The center region uses coefficients that preserve edges and object features, while the surrounding regions use coefficients that effectively reduce zipper noise. This local differentiation allows simultaneous noise reduction and feature preservation.
Solution Approach 2:
The filter pattern combines multiple filter types into a single composite filter. It integrates Gaussian filtering (for noise reduction), Laplacian filtering (for edge preservation), and sharpening operations into one unified filter pattern with specific coefficient relationships, achieving multiple functions simultaneously.
2Measurement precision
If complex filtering operations are applied to maintain object features, then stereo matching accuracy is improved, but processing complexity increases
Solution Approach 1:
Multiple filtering operations (Gaussian filtering, Laplacian filtering, sharpening) are merged into a single filter pattern application. Instead of sequentially applying multiple complex filters, the invention combines their effects into one filter pattern with predetermined coefficient relationships, reducing processing steps while maintaining effectiveness.
Solution Approach 2:
The filter pattern uses specific parameter relationships between coefficients to achieve complex filtering effects with simplified control. The coefficients follow predetermined relationships (e.g., center coefficient equals sum of surrounding coefficients, specific polarity patterns) that automatically adjust the filter behavior without requiring complex adaptive algorithms.
Data Source
AI summary
An image processing device includes a processing circuit. The processing circuit is configured to perform stereo matching based on a left image and a right image that have been subjected to filtering with the use of a filter pattern that is a pattern having four-fold rotational symmetry. The filter pattern includes: a plurality of first filter coefficients provided in a first region and having a pattern of a Gaussian filter; and a plurality of second filter coefficients having a pattern obtained by removing a part corresponding to the first region from a pattern of a sharpen filter, the second filter coefficients having values equal to each other. One of the first filter coefficients having the largest absolute value is provided at the middle of the first region. The left image and the right image are images generated by demosaicing.