研究了动态粒子图像追踪过程中的误匹配问题，提出了基于自组织映射（SOM）神经网络的粒子图像匹配算法。该方法使用SOM神经网络将归一化相关算法与最近邻判断准则结合在一起。首先使用互相关算法估计初始匹配位置；然后根据不同相关度的位置信息构建SOM神经网络并使用近邻支持判断准则选择最佳匹配位置。经SOM神经网络改进的粒子图像匹配算法大大减少了伪矢量的数量，增强了实际的处理能力；最后，使用人工合成的粒子图以及真实流场中的粒子图像进行了算法验证及误差分析。结果表明，该算法在分析精度方面有很大的提高并且具有很强的鲁棒性。

In order to reduce matching error, in this paper, a new matching method for particle images is proposed based on the SOM neural network, which combines the nearest neighbor matching algorithm with the cross correlation algorithm. Firstly, the cross correlation approach is used to evaluate the initial matching position. Secondly, the processing results of the correlation are used to build the neural network. Thirdly, nearest neighbor matching algorithm is adopted to select the best matching points. The modified method can reduce the number of false vectors and improve the practical value. At last, the synthetic particle images and real particle images are tested and the errors are analyzed. The experimental results show that the proposed method is a robust algorithm for measuring the movement of particles and the vector fields can be obtained with high precision.