超声全聚焦成像校正模型及加速算法

发布时间：2018-04-19 21:52

本文选题：超声相控阵 + 全聚焦成像　；参考：《仪器仪表学报》2017年07期

【摘要】：超声全聚焦成像算法是一种基于全矩阵数据的成像技术,具有成像精度高、缺陷表征能力强的优点,但也存在数据量大、计算时间长、近表面区域噪声较大的缺点,目前主要应用于后处理成像。针对以上问题,建立了基于阵元指向性函数的成像校正模型;结合并行计算设备,提出了基于三角矩阵数据采集和索引技术的成像加速算法。使用16阵元相控阵探头对带有人工通孔的铝制试块进行检测试验,结果表明,使用加速算法可以将每帧图像成像时间缩短至135 ms以内,满足实时成像的要求;添加校正模型后,对于图像近表面区域噪声抑制效果明显,降低了伪缺陷出现的可能性,提高了图像信噪比。
[Abstract]:Ultrasonic full focus imaging algorithm is a kind of imaging technology based on full matrix data. It has the advantages of high imaging precision and strong defect representation ability. However, it also has the disadvantages of large amount of data, long calculation time and large noise near the surface. At present, it is mainly used in post-processing imaging. Aiming at the above problems, an imaging correction model based on array element directivity function is established, and an imaging acceleration algorithm based on triangular matrix data acquisition and indexing technology is proposed in combination with parallel computing equipment. Using 16 array phased array probe to detect aluminum sample with manual through hole, the results show that the image imaging time of each frame can be shortened to less than 135 Ms by using accelerated algorithm, which can meet the requirement of real-time imaging, and after adding correction model, the image imaging time of each frame can be shortened to 135 Ms, and the correction model is added. The noise suppression effect on the near surface of the image is obvious, the possibility of false defects is reduced, and the signal-to-noise ratio of the image is improved.
【作者单位】：天津大学精密测试技术及仪器国家重点实验室;天津职业技术师范大学天津市信息传感与智能控制重点实验室;
【基金】：国家自然科学基金项目(61473205,61374219) 教育部博士点基金(20130032130001) 国家重点研发计划(2016YFF0101800)项目资助
【分类号】：TB559

【相似文献】