曲线特征匹配技术研究现状及展望

Research status and prospect of curve feature matching technology

曲线匹配在模式识别、计算机视觉和图像理解中具有重要的作用。近年来,基于特征的曲线匹配算法越来越多,主要分为基于轮廓线的匹配技术和基于纹理特征的曲线匹配技术,如Freeman链码、B样条、微分几何特征、傅里叶变换、SIFT特征、MSCD特征等。然而,相关的曲线匹配综述却几乎没有,为此,文章对近年来的主要研究进行归纳和分析,从曲线匹配的重要性、分类、优缺点等方面进行阐述,并对基于特征的曲线匹配技术的研究难点和未来趋势进行了展望。

Curve matching plays an important role in pattern recognition, computer vision and image understanding. In recent years, there are more and more feature-based curve matching algorithms, mainly the matching technology based on contour and the curve matching technology based on texture features such as Freeman chain code, B-spline, differential geometry, Fourier transform, SIFT features, MSCD features, and so on. However, the relevant curve matching review is almost no. Therefore, this paper summarizes and analyses the major researches, and states from the importance, classification, the advantages and disadvantages of curve matching. Finally, the prospect for future research and the technical difficulties of curve matching are also discussed.