图像匹配算法的研究进展

Progress on Image Matching Algorithm

图像匹配是计算机视觉和图像处理中的重要研究内容。分析了图像匹配中的难点问题及其关键技术,研究了组成匹配算法的四个要素,介绍了近来出现的新思路和新方法,对匹配算法进行了分类和性能比较,提出了实际应用中有待进一步研究的内容,如算法的集成、神经网络、遗传算法和基于高层语义的应用等。

Image matching is an important research content in computer vision and image processing. The difficulties and key techniques of image matching are analyzed, then four factors constituted image matching algorithm are studied. At last the algorithms are classified into three groups, and matching algorithms are compared with their capacities. The new idea and method proposed in recent years are presented. Some other problems such as integration of algorithms, NN, genetic algorithm and high-level semantic-based algorithms are also studied.