一种改进的曲线特征点提取方法

An improved method for extracting feature points of curve lines

针对传统曲线特征点的提取方法均采用Douglas-Peucher(D-P)算法给定阈值的思路,在曲度较大处易造成一些重要特征点缺失,在弯曲的水平距离较小处易造成非特征点保留的问题,该文提出一种改进的曲线特征点提取方法,通过自适应来提取曲线的全局特征点,即通过标准差分别计算出间隔点连线的参考值和中间点到该线垂线的参考值,比较曲线上每个点的弯曲程度与曲线整体弯曲程度的大小,来提取曲线的全局特征点。以道路和河流的实际线数据进行了实验,结果表明,该方法不仅很好地保持了曲线整体形状特征,而且大幅降低了压缩误差、提高了精度。

Aiming at the problem that Douglas-Peucher(D-P) algorithm that had been widely used for the curve feature points extraction,usually caused some important feature points to be missing in the large curvature,and leaded to non-feature point retention when the horizontal distance of bending was small. In this paper,an improved method of extracting curve characteristic points was proposed.The global feature points of the curve were extracted adaptively.The reference value of the interval point connection and the reference value from the middle point to the vertical line were calculated by standard deviation,and the bending degree of each point on the curve and the overall bending degree of the curve were compared to extract the global feature points of the curve finally.The experiment was implemented by using the actual line data of roads and rivers.The experiment results showed that the method could not only maintain the overall shape characteristics of the curve,but also greatly reduced the compression error and improved the accuracy.