摘要
提出了一种基于特征点的二维碎片轮廓匹配算法,有效地减少了匹配过程中的计算量。先对图像轮廓进行提取和去噪,计算轮廓曲线的曲率后选取特征点。接着计算特征片段的弧长与弦长比值,作为曲线特征进行匹配。实例表明,该方法提高计算的效率和匹配的成功率。
For effectively reducing the amount of computation during the matching process,a method for two-dimensional fragments matching based on feature points is presented.The main idea of this method was effectively select the feature points by calculating the curvature after extract and de-noising the contour curve of the fragment.The ratio between arch length and chord length of the feature segment were used as a key to contour matching.Experimental result shows that this method improves computing efficiency and success rate of the matching.
出处
《科学技术与工程》
北大核心
2014年第5期129-132,共4页
Science Technology and Engineering
基金
中央高校基本科研业务费专项资金(13XS01)资助
关键词
二维碎片
特征点
曲率
弧长
弦长
轮廓匹配
two-dimensional fragment
feature point
curvature
arch length
chord length
contour matching
