摘要
为了进一步提高图像匹配的速度和精度,提出SIFT结合改进的Harris的图像匹配方法,使候选点数量更少,特征点更稳定,匹配更有效率。首先用稳定的SIFT算法检测提取尺度空间极值点作为特征候选点,再下一步精确定位筛选时结合改进的Harris算法,根据灰度的"相似度"的原则进行Harris特征提取。实验结果表明,该方法大大提高了特征点提取速度和降低计算复杂度;在保持良好的匹配率的同时明显提高了算法效率和匹配速度。
In order to improve the speed and accuracy of image matching further, this paper puts forward an image matching method of SIFT combining with the improved Harris, which makes its number of candidate points less, more stable of the feature point and more efficient of matching. Firstly, the stable SIFT algorithm is used to detect and extract the extreme value point of scale space as the characteristic candidate points, the next step is to combine the improved Harris algorithm while precisely locating and screening, and to process the Harris feature extraction according to the principle of the ＂ similarity＂ of the gray. Experimental results show that this method greatly increases the speed of characteristic point extraction and reduces the computational complexity. While keeping a good matching rate, it obviously improves the algorithm efficiency and matching speed.
出处
《计算机应用与软件》
CSCD
北大核心
2013年第7期126-131,共6页
Computer Applications and Software
基金
上海市高等学校青年科学基金项目(03SQ05)
