讨论了一种基于小波变换的实时影像匹配方法。在多尺度特征提取的过程中，采用了局域正则性指数作为约束条件，保证了特征提取的可靠性；利用小波正交分解与重建实现了分层匹配；同时，采用特征与影像叠加进行了亚像素匹配。经过试验证明，上述方法是行之有效的。

A real time image matching based on wavelet transform are discussed in this paper. Lipschith regularity exponent is introduced as the constraints during multiresolution feature extraction in order to ensure the reliability. A hierarchical matchimg is realized using orthogonal wavelet decomposition and reconstruction. Sub pixel matching is carried out by superposing edge feature on image. The experimental results demonstrate that above method is significant.