摘要
为解决图像降噪过程中的弱边缘问题,提出了一种基于上下文量化和局部线性回归的医学图像降噪方法。应用信息论中上下文量化技术将复杂的混合噪声模型转换成一个局部回归分析问题,并设计了一个基于上下文的自适应滤波器;应用自适应核回归分析,进一步解决了滤波器的参数高鲁棒性估计问题;结合基于上下文的自适应滤波器技术的图像增强方法,研究了一个有效的多尺度医学图像增强算法系统。对该医学图像增强技术进行系统评估后表明,基于上下文量化的多尺度医学图像增强技术优于现有算法。
To deal with the problem of weak edges in medical image algorithm using context quantization and local linear regression. Context according to minimization of conditional entropy of the GAP prediction cells and the local texture features hidden in the context. Local signing a robust filter for pixels in each quantized context. Ex proposed algorithm outperforms conventional edge-preserving fil denoising, we present an quantization is conducted residual in the quantized linear regression is applied in de-perimental results show that the ters reviewed in this work.
出处
《上海电机学院学报》
2012年第5期325-331,共7页
Journal of Shanghai Dianji University
基金
国家科技支撑计划项目资助(2012BA11507)
国家自然科学基金项目资助(61072146)
上海市科学技术委员会浦江人才计划项目资助(10PJ1404400)
上海高校青年教师培养资助计划项目资助(shdj002)
关键词
边缘保持滤波
图像去噪
上下文量化
回归分析
edge-preserving filtering
image denoising
context quantization
regression analysis
