中子与X射线图像的WBCT域局部自适应融合算法

A local adaptive fusion algorithm for neutron and X-ray radiation images using WBCT

提出了一种基于WBCT(Wavelet-Based Contourlet Transform)的中子与X射线图像融合方法。WBCT继承了小波分析的多尺度性,并且能更有效地表达图像的高维奇异,可将不同图像的信息有效地融合在一起。同时,为了弥补基于像素的图像融合方法的不足,该算法通过邻域一致性测度的计算,实现了变换系数的局部自适应融合。实验结果表明,该方法所得的融合结果保留了更多的细节信息,提供了更多的准确信息,为中子成像和X射线成像的优势互补,提供了一种全新的研究思路。

An algorithm for neutron and X-ray image fusion using WBCT （Wavelet-Based Contouflet Transform） was proposed. The WDCT inherits the multi-scale characteristic of wavelet transform and outperforms wavelet transform in representation of high dimensional singularity of image, enabling it to synthesize information of different images effectively. In order to overcome the shortcoming of pixel-based fusion, the method realizes the local adaptive fusion through NHM （Neighborhood Homogeneity Measurement）. Experiments show that this algorithm retains more image detail and therefore provides more accurate information than traditional image fusion methods. It is proved to be a novel idea for the complementary application of neutron imaging and X-ray radiograph.