According to the shortcomings of the traditional fusion algorithm in image fusion that can ' t keep the source image information and edge information of faults well, a kind of wavelet transform image fusion algorithm based on Laplacian operator is put forward. The images are decomposed by using wavelet transform, from which the image of low frequency coefficients and high frequency coefficients are obtained. When choosing the high frequency coefficients, the fusion rule which used is based on comparison of convolution of two Laplacian operator. The high frequency coefficients use the fusion rule which is based on combination of Laplacian definition function and eight neighbourhood variance. Finally, the fused image is obtained by inverse wavelet transform. The experimental results were analyzed by the subjective evaluation and some objective evaluation standards. The result shows that the improved fusion algorithm is better than the traditional fusion algorithm on the fusion results. The fused image has the advantages of rich edge information and high image definition.%针对传统融合算法在图像融合时,不能很好地保留源图像信息和边缘信息的缺点,提出了一种基于Laplacian算子的小波变换图像融合算法.该算法利用小波变换对图像进行分解,分解后得到图像的低频系数和高频系数;对高频系数采用基于对两个Laplacian模板算子卷积结果相比较并进行筛选的融合规则;对低频系数采用基于拉普拉斯清晰度评价函数和8邻域局部方差相结合的融合规则;最后进行小波逆变换得到融合图像.对实验结果结合主观和多种客观评价方法进行分析.结果表明,该改进算法与传统融合算法相比融合效果更好,融合图像具有边缘信息丰富、 图像清晰度高等优点.
展开▼
