MRI图像的小波融合处理

Processing of wavelet fusion to MRI image

背景:小波图像融合是将两幅图像融合在一起,以获取对同一场景的更为精确、全面、可靠的图像描述。目的:用小波变换图像融合技术融合MRI脑梗死图像,以恢复缺损图像。方法:图像融合的主要机制是利用二维小波分析法对MRI脑梗死图像进行小波分解,并对高低频信号采用多种融合方式进行融合。通过对比不同融合方式后的效果图,找出最适合本部位MRI图像的融合方法。结果与结论:不同方式的融合技术能成功修复不同的缺损部位,多种融合方式的合适组合能完全修复多处缺失部位。对于文中给出的MRI脑梗死图像,采用最小值融合方式的融合效果最好。提示使用二维小波分析法处理医学图像,简便快捷,能有效改善图像的视觉效果,辅助临床诊断。

BACKGROUND： Wavelet fusion is the fusion of two or more images together to get more comprehensive and more accurate image information in same scene. OBJECTIVE： To restore the lack parts of image and improve the quality of medical image and make diagnoses more convenient. METHODS： The main principle of image fusion is to have wavelet decomposition by using two-dimensional wavelet analysis to MRI brain stem images and applies many fusion methods to approximations coefficients and details coefficients. Through comparing the renderings in different fusion methods it can identify the most suitable image fusion method for the MRI image; it is the minimum integration method. RESULTS AND CONCLUSION： The fusion methods can successfully restore the lack parts of original images. Many missing parts can be fully restored using suitable combination of all sorts effusion methods. Integration with the minimum means the best fusion to the MRI image introduced in the paper. The two-dimensional wavelet analysis of medical images can simplify the program work and improve the visual effects and be helpful to clinical diagnosis.