| 基于随机抽样一致性的多平面区域检测算法 |
| |
| 引用本文： | 赵录刚,吴成柯.基于随机抽样一致性的多平面区域检测算法[J].计算机应用,2008,28(Z2). |
| |
| 作者姓名： | 赵录刚 吴成柯 |
| |
| 作者单位： | 西安电子科技大学综合业务网理论与关键技术国家霞点实验室,西安,710071 |
| |
| 摘 要： | 在随机抽样一致性(RANSAC)的基础上,提出了一种对多个平面区域同时进行检测的算法.该算法假设对同一场景的一对未定标图像已经进行了特征点提取和匹配,首先利用对极几何约束计算出一对极点,然后随机抽取多组3对而非4对特征点定义多个待确定单应性矩阵模型,对图像对中的多个平面区域同时进行检测.模拟实验和真实实验都证明该算法具有运算量小、准确性高、鲁棒性好等优点.
|
| 关 键 词： | 多平面区域检测 单应性矩阵 随机抽样一致性 |
Planar regions detection based on RANSAC |
| |
| ZHAO Lu-gang,WU Cheng-ke.Planar regions detection based on RANSAC[J].journal of Computer Applications,2008,28(Z2). |
| |
| Authors: | ZHAO Lu-gang WU Cheng-ke |
| |
| Affiliation: | ZHAO Lu-gang,WU Cheng-ke(State Key Laboratory of Integrated Service Networks,Xidian University,Xi'an Shaanxi 710071,China) |
| |
| Abstract: | A robust method for detecting multiple planar regions simultaneity based on Random Sample Consensus(RANSAC) was proposed.Using a set of feature points that have been matched between images without calibration,a pair of epipoles were computed via homography constrain,then only 3 but not 4 pairs of feature points would be needed to define homography for confirming.Through defining multiple homographies for confirming,multiple planar regions could be detected simultaneously by using improved RANSAC.The experim... |
| |
| Keywords: | planar regions detection homography Random Sample Consensus(RANSAC) |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
