Junction and Corner Detection Through the Extraction and Analysis of Line Segments
Junction and Corner Detection Through the Extraction and Analysis of Line Segments
复制标题
DOI：
10.1007/978-3-540-30503-3_42
复制
发表时间：
2004-12
期刊：
影响因子：
--
通讯作者：
C. Perwass
中科院分区：
文献类型：
--
作者：
C. Perwass
提出了一种利用几何方法对图像边缘结构进行局部分析的算法。可被解释为拐角或交叉点的局部边缘结构被假定为可由一组线段表示。在第一步中，评估将局部边缘结构分割成线段。这导致了局部边缘结构的图模型，可以使用组合方法进一步分析该模型。结果是分类为角点或连接点，以及绝对方向和内部结构，如角点的开放角度，或连接点两腿之间的角度。给出了模拟数据和实际数据的计算结果。
An algorithm is presented that analyzes the edge structure in images locally, using a geometric approach. A local edge structure that can be interpreted as a corner or a junction is assumed to be representable by a set of line segments. In a first step a segmentation of the local edge structure into line segments is evaluated. This leads to a graph model of the local edge structure, which can be analyzed further using a combinatorial method. The result is a classification as corner or junction together with the absolute orientation and internal structure, like the opening angle of a corner, or the angles between the legs of a junction. Results on synthetic and real data are given.