A Multi-Scale Image Matching Recognition Algorithm Based on Phase Correlation

Abstract:

Article Preview

The image matching recognition method of phase correlation is based on the shift characteristics of the Fourier transform. The traditional image matching recognition algorithm has significant influence of the template size upon its matching accuracy and has weak resilience to noises except for the gauss noise. Addressing these shortcomings, we proposed a multi-scale matching recognition method based on phase correlation, combined with wavelet transform and edge detection. The algorithm, processed the reference image and the template image in different scales with such steps: decomposition, denoising, reconstruction, edge detection and the Fourier transform, phase correlation. Hence, it overcome the dependence upon template size effectively and improve the reliability and the resilience of various noises. Finally, we verified the algorithm with a real ground shooting image as the reference image and an intercepted part as the template image. The results have shown that the proposed approach is better than the traditional image matching method.

Info:

Periodical:

Advanced Materials Research (Volumes 301-303)

Edited by:

Riza Esa and Yanwen Wu

Pages:

1438-1443

DOI:

10.4028/www.scientific.net/AMR.301-303.1438

Citation:

Y. G. Tian et al., "A Multi-Scale Image Matching Recognition Algorithm Based on Phase Correlation", Advanced Materials Research, Vols. 301-303, pp. 1438-1443, 2011

Online since:

July 2011

Export:

Price:

$35.00

In order to see related information, you need to Login.

In order to see related information, you need to Login.