【摘要】 数据采集是光测力学数据处理自动化的关键和难点，特别是对条纹稀疏和低级数区．针对上述困难，本文利用ＢＰ神经网络能对函数进行逼近的特征，对条纹插值进行了探讨；经研究改进后的算法，其收敛速度明显提高．作为例子，给出了一幅光弹性等差线条纹图的插值实验，获得了令人满意的结果．更多还原

【Abstract】 Data collecting is one of the key and difficult points in automated photomechanics data processing. Especially for the fringe patterns with sparse fringes and for the areas with lower fringe orders, it forms a real difficult problem. In view of that a function can be approximated by BP neural network approach, a fringe order interpolation method based on BP neural network was explored and succeeded. The improved algorithm is much better than the traditional one in convergent speed. As an example, experiment for the interpolation of an isochromatic fringes pattern is given. The result is satisfactory.更多还原