Explore MATLAB source code curated for "描述子" with clean implementations, documentation, and examples.
Code implementation for extracting self-similarity descriptors, based on the algorithm presented in "Matching Local Self-Similarities across Images and Videos". The implementation includes key functions for computing local correlation surfaces and constructing descriptor vectors.
This code implements SIFT algorithm for image feature point extraction and descriptor computation. Running show.m performs feature point extraction, while match.m handles image matching between two images. The documentation explains SIFT algorithm workflow, with 1.jpg and 2.jpg serving as test images for demonstration.
The Scale-invariant Feature Transform (SIFT) is a computer vision algorithm for detecting and describing local image features. It identifies keypoints by finding extrema in scale space and extracts their position, scale, and rotation-invariant descriptors. Developed by David Lowe in 1999 and refined in 2004, this algorithm enables robust image matching by comparing feature descriptors. The implementation in t2.m demonstrates how to extract SIFT features and match tampered regions between two images using descriptor comparison algorithms.