基于无向图的图像整体骨架表示模型及其算法

在图像处理、模式识别领域，往往需要求解图像的骨架来获得图像的骨架来获得图像特征的最有效的数字信息。该文提出了一种基于无向图的图像整体骨架表示模型，并基于这一模型设计实现了图像的骨架化算法。它通过专门的分割算法，对图像进行分割并用无向图表示，然后对无向图中各项点采用多边形近似方法进行骨架化。算法充分利用了图像的全局和局部信息，具有速度快、效果好等优点，并得到了实际应用。文章最后给出了一些应用实例。

Undigraph-Based Whole Expression Model and Algorithm for the Skeleton of an Image

The skeleton of an image, which contains the image's most valid digital information, is often used in the areas of pattern recognition and image processing. A new skeleton model and its algorithm are introduced in this paper. The model is based on undigraph and the algorithm uses multi ploys to get the skeleton of an image. Because the algorithm makes using of not only the global information but also the local information of an image, it spends little time on skeletonisation and its affection is good. Some samples are provided in the end.