改进恒模盲均衡在医学CT图像盲恢复中的应用  

Application of improved constant module blind equalization algorithm in medical CT image blind restoration

作　　者：孙云山[1,2] 张立毅[1,2] 张锦[3] 段继忠[2] 

机构地区：[1]天津商业大学信息工程学院,天津300134 [2]天津大学电子信息工程学院,天津300072 [3]山西医科大学第一附属医院设备处,太原030023

出　　处：《仪器仪表学报》2012年第4期878-884,共7页Chinese Journal of Scientific Instrument

基　　金：天津市高校科技发展基金(20110709);山西省基础研究项目(2011011015-3)资助

摘　　要：利用图像退化与信号码间干扰产生之间的相似性,将盲均衡算法应用于医学CT(computer tomography)图像的盲恢复中。提出一种基于降维处理的分数低阶恒模盲均衡算法,算法利用线性变换将图像恢复过程等效为一维盲均衡运算,建立了降维处理的医学CT图像盲均衡恒模代价函数,算法权向量的更新中引入了代价函数的二阶Hessian矩阵,从而提高了算法的收敛速度,改善了恒模算法性能。仿真结果验证了算法的有效性,新算法改善了峰值信噪比和恢复效果,提高了运算效率。With the similarity between image degradation and inter-symbol interference,the blind equalization algorithm is applied to medical computer tomography（CT） image blind restoration.A fractional lower-order statistics image blind equalization algorithm based on dimension reduction is proposed.The algorithm adopts a linear transform to make the process of image restoration equivalent to one dimensional blind equalization.The constant modulus blind equalization cost function applied to medical CT image is established.In order to speed up the convergence of the constant modulus algorithm and improve the performance of the algorithm,the second rank Hessian matrix of the cost function is utilized in the process of weight vector update.Computer simulations demonstrate the effectiveness of the algorithm.Compared with traditional methods,the new algorithm improves the peak signal to noise ratio,restoration effect and operation efficiency.

关 键 词：盲均衡 恒模算法 CT图像 分数低阶统计量 代价函数 

分 类 号：TN911.73[电子电信—通信与信息系统]