针对离散小波变换多聚焦图像融合算法中对低频分量融合处理存在缺陷而导致图像的边缘失真及图像模糊等问题, 提出了一种自适应区域融合规则多聚焦图像融合算法。首先将源图像进行小波分解, 得到低频系数和高频系数, 然后将代表近似信息的低频系数采用改进空间频率进行阈值处理, 将代表细节信息的高频系数采用改进的梯度与改进的拉普拉斯能量和高频融合规则处理, 最后对处理后的高频系数和低频系数进行小波重构。实验结果表明: 本文算法与传统的融合方法相比, 在主观上, 图像轮廓信息和边缘信息保留更多, 融合效果更好; 在客观上, 客观指标大幅提高。

Aiming at the problem that the discrete wavelet transform multi-focus image fusion algorithm ignores the low frequency component in fusion processing and leads to edge distortion and image blur. An adaptive region fusion rule of multi-focus image fusion algorithm is proposed. First of all, the source image is decomposed by wavelet, and the low frequency and high frequency coefficients are obtained. Then the threshold processing of low frequency coefficients of the approximate information are carried out with the improved spatial frequency. Then, the improved gradient and the sum-modified-Laplacian high frequency fusion rule are utilized on the representative information. Finally, the processed high frequency and low frequency coefficients are reconstructed by wavelet transform. Experimental results show that the proposed algorithm can preserve the image contour information and edge information subjectively, and the fusion result is better than the traditional fusion methods. The objective indicators are greatly improved objectively.