• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International conference on electric and electronics;EEIC 2011 >Estimate the All Vanishing Points from a Single Image
【24h】

Estimate the All Vanishing Points from a Single Image

机译：从单个图像估计所有消失点

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Vanishing point is the important precondition for camera self-calibration from a single image. Previously proposed solutions, either relies on voting in the Gaussian sphere space, or iteratively time and again base on Maximal Likelihood Estimator. All these methods expend lots of time, have large error and low level of efficiency. Hence, a new scheme will be presented with a recently proposed algorithm called J-Linkage, in which each edge is represented with the characteristic function of its preference set and vanishing points are revealed as clusters in this conceptual space. First, it estimates all possible vanishing point, and then refines them by Expectation Maximization. Finally, two experiments show that algorithm reduces the number of variables, error measures are done in the image, a consistency measure between a vanishing point and an edge of the image can be computed in closed-form. So that it has a low computation and high precision.
机译：消失点是从单个图像进行相机自校准的重要前提。先前提出的解决方案要么依赖于高斯球空间中的投票，要么基于最大似然估计器一次又一次地迭代。所有这些方法花费大量时间，误差大且效率低。因此，将使用最近提出的称为J-Linkage的算法来提出一种新方案，其中，每个边缘都用其偏好集的特征函数表示，消失点在此概念空间中显示为聚类。首先，它估计所有可能的消失点，然后通过期望最大化对它们进行细化。最后，两个实验表明，该算法减少了变量的数量，在图像中进行了误差测量，可以以闭合形式计算消失点和图像边缘之间的一致性测量。因此它具有较低的计算量和较高的精度。

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