一种面向遥感影像自动提取结果的线要素自适应平滑方法

遥感影像提取结果的不光滑现象严重影响其数据质量和可视化效果。为解决这一问题,本文综合点、线、面要素的空间表达特征,提出了一种面向线要素的自适应平滑方法。首先,对现有的线要素平滑方法进行了总结和分析,指出了当前方法存在的局限性;然后,分析了噪声的来源和特点,提出了一种基于三角内心的自适应平滑方法;最后,以e Cognition8.0软件的道路自动提取结果为例,对方法的适用性和有效性进行了验证和分析。实验结果表明,本文提出的方法平滑效果较好,能够改善数据质量和可视化效果。

An Adaptive Line Smoothing Method for Remote Sensing Image Auto-extracting Data

The unsmoothness of the remote sensing image auto-extracted results seriously affects the data quality and visualization effect. To solve this problem, the characteristics of spatial expression of point, line and polygon element are summaries and an adaptive smoothing method for line element is proposed. First, we make a summary and analysis for the existing simplifying methods of line, and point out the limitation of these methods; then the noise source and characteristic are analyzed, and a smoothing method based on the trigonometric inner algorithm is put forward; at the end, an experiment is made to test the method proposed using the line auto-extracting data by used of eCognition8.0. The results show that the method proposed in this paper dose well for smoothing data and can improve the data quality and visualization effect for line element.