Line and point matching based on the maximum number of consecutive matching edge segment pairs for large viewpoint changing images

Luzi Wang,Yunsheng Qian,Xiangyu Kong
DOI: https://doi.org/10.1007/s11760-021-01959-6
2021-06-13
Abstract:The traditional line matching method has the problems of inaccurate endpoint location, small matching quantity and long time-consuming for images with large viewpoint variation, and the present point matching technology for the same type of image is facing the challenge of improving the matching quantity and accuracy. To solve these problems, a feature matching method based on the maximum number of consecutive matching edge segment pairs (MNCME) is proposed. Firstly, structural edges are extracted from the comparative images and segmented by edge corners. Then, the disambiguating geometric attributes are utilized to describe the edge segment, and the similarity of edge segments is represented by the sum of the differences between their geometric attributes. Finally, the edge correspondence is established based on the maximum number of consecutive matching edge segment pairs. The generated matches contain lines and points. Through combining MNCME with SIFT, the number of point matches improves greatly. To verify the performance of this method, it is compared with the state-of-the-art feature matching technology, and a series of large viewpoint changing image pairs with high-definition and abundant structural information are adopted. The results of line matching show that MNCME has better matching accuracy and time efficiency than relevant line matching technologies. The point matching results show that MNCME + SIFT is superior to other related algorithms in matching quantity, speed and correctness.
What problem does this paper attempt to address?