首页 >  2004, Vol. 8, Issue (5) : 465-470

摘要

全文摘要次数: 3337 全文下载次数: 15
引用本文:

DOI:

10.11834/jrs.20040513

收稿日期:

2003-03-04

修改日期:

2003-04-21

PDF Free   HTML   EndNote   BibTeX
实现栅格图形和图像数据矢量化提取的游程轮廓追踪法
南京大学 城市与资源学系,江苏 南京 210093
摘要:

提出一种直接利用栅格图形游程编码进行图斑边界多边形矢量化提取的有效方法。对遥感图像专题信息和栅格型地理空间分析结果的矢量化提取是GIS的重要手段 ,栅格数据结构可简化解决叠置分析等一类空间分析问题 ,其中需要高效矢量化技术的支撑。本方法充分利用栅格游程编码所具有的可操作性强、检索和转换便捷、隐含横向差异信息、可处理高精度和大规格图形和图像等特征 ,定义了“上邻游程”和“下邻游程” ,并采用相应有效查找算法和基于游程轮廓追踪的矢量化技术 ,使得计算机处理栅格数据矢量化的规模、精度和效率明显提高 ,可广泛应用于栅格型地理空间分析和遥感图像处理等方面。

A Method for Tracing Run-length Outline to Implement Vectorization of Raster Graphics and Image Data
Abstract:

An effective method of vectorization in which both raster graohics and image data are convweted into polygon features through tracing the run-length outline is developed.It is an important aspect in GIS to extract thematic vector information from remote sensing images and the resultant soatial analysis such as spatial overlay and spatial diffusion,which can be simplified in the raster data structures needs a powerful support of efficient vectorization technique.The run-length codes of raster have the characteristics of storing high precision image,containing the horizontal difference information,easy access,rapid conversion,operability and moderate compression ratio where these features are well shown in the paper.The concept of‘up neighbor run-length'and ‘down neighbor run-length'is well defined and the efficient algorithms for seeking vertical neighbor run-length for vertical tracing are developed in this paper.The adaptation of vectorization technique based on run-length outline enhances computer's ability and efficiency of handling large scale and hige precision complex raster data.The result of the origram test indicates that this method has stronger adaptability and weaker sensitivity to large scale and complex raster data.It is proved to be flexible and practical,as it can extract not only the arc links with topo-relation but also integral borders.This method could be widely used for raster-type geograohical spatial analysis and also remote sensing images processing.

本文暂时没有被引用!

欢迎关注学报微信

遥感学报交流群