Kalman滤波在辐射图像降噪处理中的应用

为改善辐射图像质量,提出了一种基于Kalman滤波的降噪算法。该方法分析了辐射图像背景噪声的特征,在合理假设其为一阶Gaussian有色噪声的基础上改写了图像的观测方程;同时,采取NSHP(non-symmetrichalfplane)模型来构造图像的过程方程。然后使用Kalman滤波算法对图像进行滤波。实验结果表明,该方法有效减弱了辐射图像中混有的噪声,和传统滤波方法相比,更好地保持了图像的细节信息,体现了自适应的优点。这表明了Kalman滤波在辐射图像降噪处理中的潜力。

Application of Kalman filtering for radiation image denoising

A denoising algorithm based on Kalman filtering was used to improve radiation image quality. The method first analyzes the characteristics of the background noise in the radiation images, and then assumes first-order Gaussian color noise to define the equivalent observation equation. Then the NSHP (non-symmetric half plane) image model is introduced to form the state equation followed by Kalman filtering. Experimental results show that as an adaptive method, the algorithm reduces the noise while retaining the image details much better than conventional algorithms. Therefore, Kalman filtering has potential for denoising radiation images.