An Image Fusion Algorithm Based on Discrete Wavelet Transform and Canny Operator - Advanced Research on Computer Education, Simulation and Modeling - page 32

An Image Fusion Algorithm Based on Discrete Wavelet
Transform and Canny Operator
Ai Deng, Jin Wu, and Shen Yang
Dept. of Information Science and Engineering,
Wuhan University of Science and Technology, Wuhan, P.R. China
[e-mail], [e-mail], [e-mail]
Abstract. Choosing one reliable and effective fusion method to determine fusion
coefficients is the key of the image fusion. This text puts forwards a new algo-
rithm based on discrete wavelet transform (DWT) and canny operator from the
perspective of the edge detection. First make original images multi-scale de-
composed using DWT, and then acquire the level, vertical as well as diagonal
edge information by detecting low-frequency and high-frequency components'
edges. Whereafter carry out a comparison of the energy of each pixel and con-
sistency verification to more accurately determine the edge points and ensure the
clarity of the fusion image. The comparison between the traditional method and
this new method is made from the three aspects: independent factors, united
factors and comprehensive evaluation. The experiment proved the usefulness of
the method, which is able to keep the edges and obtain better visual effect.
Keywords: image fusion; wavelet transform; canny operator; edge detection.
1 Introduction
Image fusion, an important branch of data fusion, aims to make a multi-level
multi-aspect processing and synthesis towards multiple source images from
multi-sensor, as a result of maximizing the complete information of the objectives and
background[1] Multi-source images include multi-sensor images, multi-source remote
sensing images, multi-focus images and time series (dynamic) images. This paper
mainly analyzes multi-focus images fusion. It's very difficult to get a completely clear
image for the identical optical sensor resulting from the restriction of the depth of the
field when sensors image the same scene[2]. Multi-focus images have different clear
domain and fuzzy domain. The fusion goal lies to synthesize all clear domain of all
original images in order to gain a fusion image with optimum clarity.
Recently, fusion algorithms based on multi-resolution decomposition have been
extensively studied. Multi-resolution decomposition enables to decompose a image
into several images with a progressive lower resolution, that is the decomposition can
generate a set of low-pass or band-pass images. This hierarchical structure makes full
use of the local information and the global information of images[3]. In 1983, Lapla-
cian pyramid transform was proposed by Burt P.J. and Adelson E. H., which was the
first multi-resolution fusion algorithm[4]. Subsequently, contrast pyramid transform,
morphological pyramid transform and gradient pyramid transform were all used in