Abstract: Chinese patches appearance is one of the key indexes for quality.Image registration based on the combination of feature-based matching and area-based matching is used here for Chinese patch images registration.This algorithm does Harris corner detection，then standard Fourier-Mellin transform for further registration on the Chinese patches，re-sample image fusion technique is finally used for realizing Chinese patch images registration perfectly.It is used in the image detector above the Chinese patches.Chinese patches marked with drapes of different sizes and shapes are employed to detect the precision of this algorithm registration.Results show that the algorithm is not only reasonable，but also guarantees the high accuracy of Chinese patch images registration.

Key words: image registration, Chinese patch, Harris corner, Fourier-Mellin algorithm, re-sampling

摘要： 中药贴剂外观是中药贴剂质量的重要指标。把基于特征与基于区域的图像配准算法结合运用，进行中药贴剂图像配准。该方法将贴剂图样进行Harris角点检测，通过标准Fourier-Mellin匹配方法进行校验，重采样的图像融合，实现对检测过程中采集到的中药贴剂图像进行无缝拼接，应用于中药贴剂上方图像检测仪。实验用不同大小、不同形状的褶皱对中药贴剂外观表面进行标记，以便检测该算法配准精度。结果表明，该算法应用合理，能够保证中药图像配准的高精度要求。

关键词: 图像配准, 中药贴剂, Harris角点, Fourier-Mellin算法, 重采样