电子散斑干涉技术(ESPI)中,基于偏微分方程(PDE)的滤波模型是一种重要的滤波方法。偏微分方程滤波模型中的微分算符通常利用差分近似表示。给出了中心差分、九点差分、高阶差分三种不同的差分格式。以典型有效的方向二阶偏微分方程滤波模型为例,分别利用三种不同的差分格式近似滤波模型中的微分算符,通过模拟条纹图、相位图以及实验条纹图进行了分析研究,结果表明,对于密度变化特别大的条纹图,采用高阶差分格式能够更好地平衡高密度区域和稀疏密度区域的滤波效果,九点差分和中心差分格式需要使用均值滤波做进一步的处理,中心差分格式处理速度最快,高阶差分格式次之,九点差分格式则最慢。 Electronic speckle interferometry (ESPI), based on partial differential equation (PDE) filter model is an important filtering method. Differential operators in partial differential equations are usually represented using differential approximation. Three different difference schemes of center difference, nine difference and high order difference are given. Taking the typical effective second-order partial differential equation filter model as an example, differential operators in the filter model are approximated by three different difference schemes respectively. The simulation fringe pattern, phase diagram and experimental fringe pattern are analyzed and studied. The results show that , The stripe pattern with particularly large density changes adopts the high-order difference scheme to better balance the filtering effects in high-density areas and sparse density areas. The nine-point difference and center difference formats need to be further processed by averaging filtering. The central difference format The fastest processing, high-end differential format followed, nine-point differential format is the slowest.