基于高阶加权和迷宫算法的层析成像射线追踪法  

Tomography Ray-tracing Based on High-order Weighting and Maze Algorithm

作　　者：黄晓寒[1] 王仲刚[1] 袁野[1] 张彧[1] 

机构地区：[1]后勤工程学院军事土木工程系,重庆401311

出　　处：《无损检测》2017年第2期34-37,47,共5页Nondestructive Testing

摘　　要：通过对LTI线性插值法的追踪公式和全局扫描方法的改进,提出了一种基于高阶加权和迷宫算法的层析成像射线追踪法(HLTI)。该算法通过引入Taylor加权高阶项对LTI追踪公式进行改进,减小了由于线性条件假设而产生的累积误差,提高了射线追踪精度,解决了LTI追踪公式出现复值解而失效的问题;并利用迷宫算法原理对全局扫描方式进行优化,解决了传统扫描方法不能反向和竖向追踪的问题。仿真试验结果表明,新算法HLTI较传统的LTI算法具有更高的射线追踪精度和更好的成像效果。Through the improvement to the LTI track formula （linear travel-time interpolation） and global scanning method, a HLTI based on higher order weighting and maze algorithm has been advanced. By means of improving LTI track formula via bringing in Taylor higher order weighted item, this new algorithm reduces the accumulative error caused by linear condition hypothesis, improves the accuracy of ray-tracing, and thus solves the invalidation problem come about by the complex-value solution of LTI track formula. Besides, by taking advantage of the maze algorithm principle to optimize the global scanning method, the problem of the traditional scanning method which is incapable of backward and vertical tracing has been solved. The simulation experiment results indicate that the new HLTI has higher ray-tracing accuracy and better image quality than the traditional LTI.

关 键 词：LTI线性插值法 泰勒展式 CT技术 扫描优化 

分 类 号：TG115.28[金属学及工艺—物理冶金]