一种快速提高图像特征匹配算法

In a different environment of background, due to the affection of illumination, scale, etc, the feature matching of the same object caused a lot of affection. Using classical SIFT algorithm is effective overcome the affection of illumination, rotation, scale,etc. In order to improve the effective of feature matching, this paper adopt an improved SIFT algorithm. The improved way is that creating feature descriptor with constraint conditions in generate the concentric circles and adapt bidirectional matching in the basis of analysis original classical algorithm. Through the experiment proved. In adapting the concentric circles to generate feature descriptor matching to effective reduce the matching time and improve matching accuracy in guarantee algorithm stability at the same time. Effectively improve the efficiency of feature matching.?????