【摘要】 目的利用非局部均值滤波与小波相结合的算法,抑制脑部CT图像噪声,提高图像的质量。方法通过仿真实验确定不同噪声水平下的滤波系数,然后采用真实的含噪脑CT图像进行验证,并与传统非局部均值滤波进行比较。最后采用配对t检验,对该方法滤波前后的图像峰值信噪比进行统计学分析。结果该方法能够使含有白噪声的CT图像的峰值信噪比提高5～10 dB,高出传统非局部均值滤波后图像3～5 dB。滤波前后的图像峰值信噪比具有统计学差异(P<0.001)。结论结合小波的自适应非局部均值滤波可以对不同噪声水平的脑CT图像进行自适应处理,有效去除噪声,提高峰值信噪比,同时保留图像的细节及边缘,改善图像质量。更多还原

【Abstract】 Objective To suppress the noise and improve the quality of brain CT image by the method based on a combination of non-local mean filtering and wavelet. Methods Experiments were conducted on simulation data to estimate the filtering parameters under different noise levels and the brain CT image from human subjects to demonstrate the validity of the proposed theory. The effect was compared with the traditional non-local mean filtering. Finally, the peak signal to noise ratio(PSNR) of the image before and after filtering was statistically analyzed by paired t test. Results This method can increase the PSNR of the CT image with white noise by 5～10 dB, which was 3～5 dB higher than the traditional non-local mean filtered image. The PSNR of the image before and after filtering was statistically different(P<0.001). Conclusion The method based on a combination of non-local mean filtering and wavelet can process the brain CT images with different noise levels adaptively, effectively suppress noise, improve PSBR, while retaining image details and edges, and improving image quality.更多还原