关键词: 铁路铸件, 射线检测, 小波变换, 有限脊波变换, 图像融合, 图像增强

Abstract: A radiographic image enhancement algorithm based on finite ridgelet transform is proposed.In terms of characteristic of the collected radiographic signals，in the proposed scheme the subsection of signals is mapped to 0-255 gray scale to form several gray images and then fuses these images to form a new enhanced image.In the fusing process，neighboring relativity and regional variance analysis method is adopted as the low-frequency image coefficients fusion rule.High-frequency image fusion rules are based on maximum absolute value of ridgelet coefficients.All the features and details from different gray images are fused successfully and four quantitative fusion performance criteria are used to evaluate the gain.Experimental results show that the proposed algorithm can effectively elevate image’s definition and retains more subtle image details than the method using wavelet.

Key words: railway casting, radiographic testing, wavelet transform, finite ridgelet transform, image fusion, image enhancement