匹配点分布密度约束下的基础矩阵估计

来源 :武汉大学学报(信息科学版) | 被引量 : 25次 | 上传用户:whl98122368
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
提出一种匹配点分布密度约束下的基础矩阵估计方法。该方法以传统RANSAC方法为基本框架,结合匹配点分布密度约束来选择内点集,并采用M-Estimators方法重新计算基础矩阵。通过模拟数据和真实图像实验表明,本文方法可有效提高基础矩阵的计算精度。
其他文献
Sensitivity analysis of thermal equilibrium parameters in the reservoir module of MIKE 11 model was conducted for the Wuxikou Reservoir in Jiangxi Province of China in order to apply the module to the
期刊
During the last 30 years,China has witnessed rapid economic growth and dramatic urbanization,with about 1.2 × 107rural people migrating annually into urban areas.Meanwhile,especially since 1995,the ru
期刊
China is experiencing a fast process of farmland conversion which is conducted mainly by local governments to fuel economic development.Social tension is aroused and policy makers take great concerns
期刊
Understanding the Changing Planet: Strategic Directions for the Geographical Sciences(UCP),published by the National Research Council(USA),and its comments in The Professional Geographer propose that
期刊
Land use/cover change is an important parameter in the climate and ecological simulations. Although they had been widely used in the community,SAGE dataset and HYDE dataset,the two representative glob
期刊
It is believed that the global CO2emissions have to begin dropping in the near future to limit the temperature increase within 2 degrees by 2100. So it is of great concern to environmentalists and nat
期刊
Understanding the driving forces and mechanism of land use change is a key issue in land change science,and has received much attention over the past 30 years. While many driving forces have been iden
期刊
As one of the most important metropolitan areas in China,Xian City plays a leading role in the development of the western regions. To impede the decision and policy preferences for environmental and e
期刊
提出一个基于基因优化产生Tuned模板的新方法——基因扰动优化法(GDO法)。首先在初始模板群体中,根据求得的适应度值较好的少数几个模板,确定为优秀模板;然后对优秀模板中的每个基因分别进行扰动,根据扰动前后适应度值增加(或减少),确定每个优秀模板的两个优秀基因;最后,对优秀基因的数值引入扰动,得到一组新模板,进入下一代的计算,如此不断地优化得到最佳Tuned模板。经理论分析和对真实图像分类的实验证
期刊
针对启发式云变换只能对一维特征进行变换的缺陷,提出一种基于高斯混合模型的改进云变换方法。通过EM算法和高斯分布的拟合误差求解云模型的数字特征,抽取图像底层概念,从而实现图像分割。通过图像分割实验验证了该算法的有效性。
期刊