A Feature Points Matching Algorithm Based on Wavelet Analysis

摘要

In this paper we present an application of wavelet analysis on feature points matching. Wavelet coefficients which produced by wavelet transform can best display part character of the image. A new method based on the weighted statistics of wavelet coefficients for feature points matching is proposed. Through the new feature description and feature matching method, we can build a feature point database for each image and improve the efficiency of feature points matching. The experimental results show that this method is practicability, feasibility, good precision and fast speed.