A novel image matching method based on SIFT

作者：Lin Yuan Sheng; Xu Gang^*; Jiang Ming; Zhou Peng; Jiang Juan Juan

来源：Sensors and Transducers, 2014, 171(5): 276-282.

• Cosine similarity
• Feature
• Image processing system
• Keypoints
• Matching methods
• Scale invariant feature transform algorithms (SIFT)
• SIFT
• Strong robustness

The research on image matching method has been one of the main research focuses in recent years. In order to reduce the number of incorrect matching points and improve the performance of image matching in the image processing system, this paper presents a novel image matching method based on Scale Invariant Feature Transform (SIFT) algorithm and cosine similarity. The method extracts the SIFT keypoints of image1 and image2 firstly, and then makes the image matching by computing the cosine similarity of the extracted SIFT feature vectors. Extensive experimental results on the Middlebury Stereo Datasets show that the novel image matching method reduces the error matching and improves the matching effectively, with strong robustness and a better matching effect.

• 出版日期2014
• 单位Polytechnic university