基于方向可控金字塔的图像融合算法

为克服传统正交小波变换在进行图像融合时存在的不足，提出了一种基于方向可控金字塔的图像融合算法。首先对待融合图像进行方向可控金字塔分解，对分解后的低频分量采用平均和选择相结合的方法进行融合，对各方向的高频分量则使用像素绝对值选大的规则进行融合，最后对融合后的低频分量和高频分量进行方向可控金字塔重构得到融合图像。仿真试验表明算法能够得到质量较高的融合图像，同时，熵、平均梯度和空间频率等客观评价指标也较平均法和基于小波变换的图像融合算法有所提高，是一种有效的图像融合算法。

An Image Fusion Algorithm Based on Steerable Pyramid

To overcome the shortcoming of wavelet transform in image fusion, a image fusion algorithm based on steerable pyramid is proposed in this paper. The original images are decomposed by steerable filters firstly and we can get the low- frequency coefficients and high- frequency coefficients. The average and selection fusion rule is used for low- frequency coefficients. The fused high- frequency coefficients are obtained by comparing the max absolute value he clarity of low- frequency sub- image is et firstly and we can get the lows of the pixels. The fused image can be achieved by using steerable pyramid reconstruction on the fused low- frequency coefficients and high- frequency coefficients. Computer simulation results demonstrate the proposed algorithm can get better image. The objective evaluation indexes value such as Shannon, average gradient and space frequency are improved comparing with the average method and the traditional wavelet algorithm. The algorithm is effective for the image fusion.