Adaptive Chessboard Corner Point Detection via Homography Expansion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting chessboard corner points are incomplete in covering the camera's field of view and fail to accurately extract feature points, especially in complex illumination conditions or low-resolution images with large distortions, leading to inaccurate camera calibration.
Innovation Solution
A method for adaptively detecting sub-pixel level chessboard corner points involves setting marks on an initial unit grid, calculating pixel coordinates, using a homography matrix to expand detection, and dynamically adjusting the iteration window size to ensure accurate detection across the entire chessboard region.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing corner point detection methods are used, then the detection process is simple, but the coverage of camera field of view is incomplete and calibration accuracy is insufficient
Solution Approach 1:
The chessboard is divided into multiple unit grids, with the first unit grid serving as a reference and subsequent unit grids being detected through homography transformation. This segmentation allows complete coverage of the camera field of view while maintaining detection accuracy through systematic processing of each grid unit.
Solution Approach 2:
The first unit grid is detected and processed preliminarily to establish a reference framework. The homography matrix is pre-calculated based on the first unit grid's corner points, which then enables efficient detection of subsequent unit grids without requiring separate processing for each, thus improving overall accuracy while controlling complexity.
2Measurement precision
If fixed iteration window size is used for sub-pixel level corner point detection, then the processing is fast, but detection accuracy deteriorates in low-resolution or large distortion images
Solution Approach 1:
The iteration window size is dynamically adjusted based on the detected unit grid's resolution and distortion characteristics. For low-resolution or highly distorted images, the window size is enlarged to maintain detection accuracy, while for standard images, a smaller fixed window size is used to preserve processing speed. This dynamic adaptation resolves the contradiction between accuracy and productivity.
Solution Approach 2:
The detection parameters, specifically the iteration window size, are changed adaptively based on image characteristics. By modifying this key parameter according to the unit grid's properties, the system achieves optimal detection accuracy across varying conditions without significantly compromising processing efficiency.
3Measurement precision
If manual marking of initial unit grid is required, then detection accuracy can be improved, but operation complexity increases
Solution Approach 1:
The system automatically detects corner points and calculates homography matrices without requiring manual intervention for most operations. The only manual step is marking the first unit grid, which then enables automatic detection of subsequent grids through algorithmic processing, thus maintaining high accuracy while minimizing operational complexity.
Solution Approach 2:
Manual marking is required only for the first unit grid as a preliminary action to establish the reference framework. Once this initial marking is done, the system automatically processes subsequent unit grids using homography transformation, significantly reducing the operational burden while maintaining detection accuracy.
Data Source
AI summary
The present invention discloses a method for adaptively detecting chessboard sub-pixel level corner points. Adaptive detection of chessboard sub-pixel level corner points is completed by marking position of an initial unit grid on a chessboard, using a homography matrix H calculated by pixel coordinates of four corner points of the initial unit grid in a pixel coordinate system and world coordinates in a world coordinate system to expand outwards, adaptively adjusting size of an iteration window in the process of expanding outwards, and finally spreading to the whole chessboard region.