摘要
目的:医学影像在获取、存储、传输过程中会不同程度地受到噪声污染,这极大影像了其在临床诊疗中的应用。为了有效地滤除医学影像噪声,提出了一种混合滤波算法。方法:该算法首先将含有高斯和椒盐噪声的图像进行形态学开运算,然后对开运算后的图像进行二维小波分解,得到高频和低频小波分解系数。保留低频系数不变,将高频系数经过维纳滤波器进行滤波,最后进行小波系数重构。结果:采用该混合滤波算法、小波阈值去噪、中值滤波、维纳滤波分别对含有混合噪声的医学影像分别进行滤除噪声处理,该滤波算法去噪后影像的PSNR值明显高于其他三种方法。结论:该混合滤波算法是一种较为有效的医学影像噪声滤除方法。
Objective： Medical image in acquisition, storage, and transport process may be polluted by all kinds of noise in different degree. In order to effectively filter medical image noise, a hybrid filter algorithm was presented. Methods： Firstly, the medical image which contains Gaussian and salt ＆ pepper noise was implemented morphological opening operation. Then medical image through two-dimensional wavelet decomposition, retaining the low frequency wavelet coefficient unchanged, at the same time the high frequency wavelet coefficient was conducted wiener filtering.Finally, the wavelet coefficient were reconstructed. Results： Using this hybrid algorithm, wavelet shareholding denoising, median filtering, and wiener filters respectively on the medical image which contains mixture noise filtering processing. The algorithm is significantly higher than the PSNR value to the other three kings of filtering methods. Conclusion： The hybrid filter algorithm is a kind of relatively effectively medical image noise fitter method.
出处
《现代生物医学进展》
CAS
2011年第20期3954-3957,共4页
Progress in Modern Biomedicine
关键词
小波阈值去噪
数学形态学
维纳滤波
混合滤波算法
Wavelet shareholding denoising
Mathematical morphology
Wiener filters
Hybrid filtering algorithm
