Fusion Evaluation of X-Ray Backscatter Image and Holographic Subsurface Radar Image

X-ray backscatter imaging technology and Holographic Subsurface Radar can detect surface invisible targets nondestructively. These two techniques differ in the imaging mechanisms and the target characteristics they reflect, therefore, the existence of target and its electromagnetic characteristics can be known by merging these two images. In this vein, a number of fusion methods based on Weighted Fusion algorithm, Wavelet transform (WT), Non-Subsampled Contourlet transform (NSCT), Guided Filtering Fusion (GFF) and Shearlet transform are evaluated in this paper. Non-local mean filter and principal component analysis (PCA) are applied to X-ray backscattered images, and the classical Otsu method is used to extract target automatically. Experiment results from image quality metric clearly demonstrate the better performance of Shearlet transform fusion in terms of less false information, better color fidelity and better image readability.