GUO Bingxuan, LI Deren, LEI Zhen. Numeric Annotation Abstraction from Map Based on Contour Shape Analysis[J]. Geomatics and Information Science of Wuhan University, 2001, 26(3): 275-278.
Citation: GUO Bingxuan, LI Deren, LEI Zhen. Numeric Annotation Abstraction from Map Based on Contour Shape Analysis[J]. Geomatics and Information Science of Wuhan University, 2001, 26(3): 275-278.

Numeric Annotation Abstraction from Map Based on Contour Shape Analysis

  • A method for abstracting the numeric annotations from topographic map according to contour shape analysis is presented.Because the contour shape concludes full shape information of a figure,it is possible to abstract annotations according to it avoiding aberration caused by skeleton abstracting methods. In this article,we described a five-step process for numeric annotation abstraction and practical arithmetic for each step:1) Get contour shape pixels from map.2) Trace the pixels and connect them in the form of link-code.3) By improved Dichotomy vector generating arithmetic,express the link-code contour shape into vector form.4) Decompose the contour shape vector into segments.5) Abstract annotation from char segments. As emphasis,we described the arithmetic of contour shape expression and decomposition.We presented a contour shape decomposing arithmetic according to the complex degree calculate from curvature and one time difference of contour shape,which can distinguish character segment from general segment.Because different figures have different complex degrees,by defining and calculating the complex degrees of each contour shape,we divided the contour shape lines segments into two classes:character segments and general line segments.We used such division to abstract the numeric annotations from topographic maps.We also presented an improved Dichotomy vector expression,which can balance the quality and cost requirement. At the end of this article,we presented the result of experiment.It proved that the process and arithmetic described above is practical and efficient.
  • loading

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return