计算机工程与应用
計算機工程與應用
계산궤공정여응용
COMPUTER ENGINEERING AND APPLICATIONS
2013年
10期
188-190
,共3页
加速稳健特征(SURF)重构%特征提取%直方图均衡化
加速穩健特徵(SURF)重構%特徵提取%直方圖均衡化
가속은건특정(SURF)중구%특정제취%직방도균형화
Speeded Up Robust Feature(SURF)reconstruction%feature extraction%histogram equalization
针对SURF算法提取图像特征点较少的问题,提出了一种经直方图均衡化处理,重构SURF尺度空间(R-SURF)的图像特征提取方法.该方法能提高图像特征点检测数目,同时保持较高的匹配率,并且继承SURF算法的良好特性.将算法与SURF和C-SURF算法进行比较实验,结果表明R-SURF具有更好的特征检测能力.
針對SURF算法提取圖像特徵點較少的問題,提齣瞭一種經直方圖均衡化處理,重構SURF呎度空間(R-SURF)的圖像特徵提取方法.該方法能提高圖像特徵點檢測數目,同時保持較高的匹配率,併且繼承SURF算法的良好特性.將算法與SURF和C-SURF算法進行比較實驗,結果錶明R-SURF具有更好的特徵檢測能力.
침대SURF산법제취도상특정점교소적문제,제출료일충경직방도균형화처리,중구SURF척도공간(R-SURF)적도상특정제취방법.해방법능제고도상특정점검측수목,동시보지교고적필배솔,병차계승SURF산법적량호특성.장산법여SURF화C-SURF산법진행비교실험,결과표명R-SURF구유경호적특정검측능력.
In order to solve the problem of less features extraction of SURF, an image characteristics extraction algorithm, R-SURF, which is based on histogram equalization and SURF reconstruction is proposed in this paper. This algorithm can extract more features while maintaining good repeatability, and keeping the robust of SURF. Experimental results show that R-SURF has a better performance than that of SURF and C-SURF.
