The Novel Method of Moving Target Tracking Eyes Location based on SIFT Feature Matching and Gabor Wavelet Algorithm

Abstract

SIFT is a scale and rotation invariant feature point extraction algorithm, and it has a high robustness of local feature representation. SIFT expression method is based on the detection of multi-scale extreme points of the difference of Gaussian image pyramid and gradient direction. This article describes a face based on Gabor wavelet transform facial area, and it is depression terrain feature points extracted directly from the gray-scale image. The paper presents the novel method of Moving target tracking Eyes Location based on SIFT feature matching and Gabor wavelet algorithm. This method is not sensitive to noise, in real-time positioning and accuracy has greatly improved. Simulation results validate the effectiveness of the proposed algorithm.

Keywords

• Gabor wavelet
• Moving object tracking
• Scale-invariant feature transform (SIFT).