...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>International Journal of Innovative Computing Information and Control >A MIRROR-CHECKING ALGORITHM FOR EFFICIENT CIRCLE/ARC DETECTION
【24h】

A MIRROR-CHECKING ALGORITHM FOR EFFICIENT CIRCLE/ARC DETECTION

机译：高效检测圆弧/圆弧的镜面检查算法

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The iterative multi-step scheme has been applied to circle/arc detection. It usually includes three main steps: picking initial points, finding correspondent searching points with predefined geometric properties, and obtaining candidate circles/arcs. A successful iteration for finding a candidate circle/arc depends on picking valid initial points and finding valid correspondent searching points, and the "valid" means those points must lie on the same target circles/arcs. That is, the iteration is redundant if initial points or searching points are invalid. In this paper, an efficient circle/arc detection method (MCD) based on a mirror-checking algorithm is proposed: we first randomly pick two initial points and construct two corresponding patterns by collecting neighbor points around the initial points. Then, we use the proposed mirror-checking algorithm to check whether the two patterns are mirrored. If they are not mirrored, we will ignore the initial points and find new ones to avoid redundant iteration; otherwise, a candidate circle can be obtained from the mirrored patterns. Instead of finding searching points within a predefined geometric region, the searching point is picked from the mirrored patterns. Based on the initial and searching points, the verification process is utilized to confirm whether the candidate circle is the target circle. From the experimental results, the proposed algorithm can efficiently reduce redundant iterations and executing time. In the meanwhile, we plug the proposed mirror-checking algorithm to the effective voting method (MEVM) and semi-random detection method (MSRD). The experimental results show that it can speed up the two methods and sift redundant iterations efficiently.
机译：迭代多步方案已应用于圆/弧检测。它通常包括三个主要步骤：选取初始点，查找具有预定义几何属性的对应搜索点以及获取候选圆/弧。查找候选圆/弧的成功迭代取决于选择有效的初始点并找到有效的对应搜索点，并且“有效”表示这些点必须位于相同的目标圆/弧上。即，如果初始点或搜索点无效，则该迭代是多余的。本文提出了一种基于镜像检查算法的高效圆弧检测方法：首先随机选取两个初始点，然后通过收集初始点周围的相邻点来构造两个相应的模式。然后，我们使用提出的镜像检查算法来检查两个模式是否镜像。如果它们没有被镜像，我们将忽略初始点并找到新的点，以避免重复迭代；否则，可以从镜像图案中获得候选圆。代替在预定的几何区域内找到搜索点，而是从镜像图案中选择搜索点。基于初始点和搜索点，使用验证过程来确认候选圆是否为目标圆。从实验结果来看，该算法可以有效减少冗余迭代和执行时间。同时，我们将提出的镜像检查算法插入到有效投票方法（MEVM）和半随机检测方法（MSRD）中。实验结果表明，它可以加快两种方法的速度，并有效地筛选出多余的迭代。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号