Explore MATLAB source code curated for "配准" with clean implementations, documentation, and examples.
This MATLAB implementation of the ICP algorithm reads 3D point cloud data and features duplicate registration avoidance, robust ICP registration with outliers, and manual outlier removal for improved point cloud alignment.
This program reads an input image and extracts SIFT features, visualizing the feature points on the original image. It also performs feature matching and registration between two images using keypoint detection and descriptor comparison algorithms.
This algorithm performs global translation registration between any two images using similarity metrics and optimization techniques, supporting diverse image types including natural, medical, and satellite imagery.
This MATLAB program utilizes the SUSAN operator to detect corner points in images, computes 128-dimensional feature vectors at these corner locations, matches feature vectors between reference and real-time images, and achieves precise image registration to obtain affine transformation parameters.
A comprehensive guide to image registration using Fourier-Mellin transform, including implementation code descriptions, preprocessing techniques, and post-processing optimization methods
ICP Algorithm - Iterative Closest Point method primarily used for 3D point cloud registration, an iterative nearest point algorithm that converges to local minima through transformation optimization
Registration of Two Images Using Mutual Information as a Similarity Measure
This program implements infrared and visible image registration by utilizing edge information for image matching, performing edge detection to identify closed contours, and using chain code representation for precise image alignment.
MATLAB-based image registration utilizing mutual information for accurate alignment of medical, remote sensing, or computer vision images
A robust non-rigid registration program capable of handling local deformations, featuring integrated registration algorithms with both reference and target images included. The implementation incorporates deformation field computation and optimization techniques for precise feature alignment.