基于全局颜色传递的SIFT匹配算法

SIFT matching algorithm based on global color transfer

研究在不同光照条件下两幅彩色图像的匹配问题,提出了一种新的基于全局颜色传递的具有尺度不变性的特征变换(SIFT)匹配算法。新算法对不同光照下同一场景或目标的两幅彩色图像进行全局颜色传递,以减小匹配时由颜色差异带来的误差;利用SIFT算法提取处理后的图像的特征信息完成初步匹配;采用随机抽验一致性(RANSAC)算法消除误匹配点。实验结果表明新算法具有良好的彩色图像匹配性能。

This paper investigates the matching problem between two color images coming from the same scene under different illuminations, and proposes a new image matching algorithm with the Scale Invariant Feature Transform（SIFT）based on global color transfer. Firstly, the new algorithm implements the global color transfer between two color images which have different illumination under the same scene, resulting in reducing the error caused by color difference. Then,the SIFT algorithm is used to extract the features information from the processed images and the primary matching between the source image and the processed one is finished. Lastly, the false matched features are discarded by using the Random Sample Consensus（RANSAC）algorithm. Simulation results indicate that the new method has better performance of color images’ matching than that in the literatures.