【摘要】 当原始相位超过周期π时,定量相衬成像中多峰相位分布与包裹相位会出现增减交替,存在多极值点.针对该问题,提出一种基于自适应阈值翻转的相位解包裹算法(SATR),实现了定量相衬成像中的真实相位重构,分析了相衬成像原理和自适应阈值翻转相位解包裹算法的实现过程.通过模拟验证计算,并与最小二乘法(LS)、横向剪切最小二乘法(LSBLS)以及四向横向剪切最小二乘法(FLSBLS)的结果进行对比,证明该算法的可行性与准确性.该算法在一定程度上提高了相位解包裹的精确度,为强干扰、高精度及大计算量的解包裹提供了新的思路.更多还原

【Abstract】 To solve the problems in quantitative phase contrast imaging such as the alternate increase and decrease between multi-peak phase distribution and wrapping phase distribution and the existence of multiple extremum points of the wrapping phase when the original phase exceeds the period π,a phase-unwrapping algorithm based on self-adaptive threshold reversal(SATR) is proposed.The algorithm realizes real phase reconstruction in quantitative phase contrast imaging.The realization of the principle of phase contrast imaging and the self-adaptive threshold reversal algorithm is analyzed,and the feasibility and accuracy of the method are proved through simulation verification calculation and comparison with Least-Square phase unwrapping algorithm(LS),Least-Square algorithm based on Lateral sheering interferometry(LSBLS) and Four-Direction Least-Square phase unwrapping algorithm based on Lateral shearing interferometry(FLSBLS).The algorithm improves the precision of phase unwrapping to some extent and provides a new idea for phase unwrapping under strong interference,demand for high precision and a large amount of calculation.更多还原