基于改进Hausdorff距离的图形注记匹配方法

Graphic Annotation Matching Method Based on Modified Hausdorff Distance

摘要:

AutoCAD 作为常用的计算机辅助制图软件，在工程领域被广泛应用。在实际绘图任务中，常常需要高效且准确地完成图形对象与人工注记文本信息之间的匹配。该匹配问题可抽象为2个集合对象之间距离和最短的组合优化问题。尽管通常采用动态规划、贪心算法、最近邻等构造性算法进行求解，但仍面临着算法复杂度较高的问题。为此，提出了一种基于改进Hausdorff距离的模拟退火算法，利用改进后的Hausdorff距离计算方法，并结合模拟退火算法的概率搜索特性，以获得匹配问题的近似最优解。通过与一系列算法的对比试验，验证了所提方法的有效性和准确性。试验结果显示，该算法能够明显提高图文匹配的精度和计算效率，适用于实际的工程制图任务。

Abstract:

AutoCAD, as a commonly used computer-aided drafting software tool, is widely applied in engineering fields. In practical drafting tasks, it is often necessary to efficiently and accurately match graphical objects with manually annotated textual information. This matching problem can be abstracted as a combinatorial optimization problem of the shortest distance between two sets of objects. Although dynamic programming, greedy algorithms, and nearest neighbor methods are commonly used to solve such problems, the problem of high algorithmic complexity is still faced. Therefore, a simulated annealing algorithm based on a modified Hausdorff distance is proposed. By employing an enhanced calculation method of Hausdorff distance combined with the probabilistic search characteristics of the simulated annealing algorithm, an approximate optimal solution to the matching problem is obtained. Through a series of comparative algorithm experiments, the effectiveness and accuracy of the proposed method are verified. The experimental results indicate that the algorithm can significantly improve the precision and computational efficiency of graphic-text matching, which is suitable for the practical engineering drafting tasks.