为了实现输电塔远距离位移监测,同时满足低成本、无接触、易实施及精确度高等要求,结合输电塔的内轮廓特征和计算机视觉位移识别技术,提出感兴趣区域(region of interest,简称ROI)关键点法。首先,利用N近邻最小能量法进行ROI轮廓搜索提...为了实现输电塔远距离位移监测,同时满足低成本、无接触、易实施及精确度高等要求,结合输电塔的内轮廓特征和计算机视觉位移识别技术,提出感兴趣区域(region of interest,简称ROI)关键点法。首先,利用N近邻最小能量法进行ROI轮廓搜索提取,并与Harris角点检测算法相结合;其次,通过输电塔台架实验与灰度模板匹配法相比,ROI关键点法位移识别结果的平均误差、均方根误差分别降低了56%和45%,绝对误差小于5 mm和10 mm的准确率提高了61%和3%,计算效率提高了11倍,稳定性及抗噪性能较高;最后,在实验塔对比验证中,ROI关键点法的位移测量值与实际位移的差值百分比在0.0%~11.1%之间。结果表明,ROI关键点法在输电塔结构位移监测中具有较高的准确率、精细度、计算效率、稳定性及鲁棒性。展开更多
Bagging is not quite suitable for stable classifiers such as nearest neighbor classifiers due to the lack of diversity and it is difficult to be directly applied to face recognition as well due to the small sample si...Bagging is not quite suitable for stable classifiers such as nearest neighbor classifiers due to the lack of diversity and it is difficult to be directly applied to face recognition as well due to the small sample size （SSS） property of face recognition. To solve the two problems,local Bagging （L-Bagging） is proposed to simultaneously make Bagging apply to both nearest neighbor classifiers and face recognition. The major difference between L-Bagging and Bagging is that L-Bagging performs the bootstrap sampling on each local region partitioned from the original face image rather than the whole face image. Since the dimensionality of local region is usually far less than the number of samples and the component classifiers are constructed just in different local regions,L-Bagging deals with SSS problem and generates more diverse component classifiers. Experimental results on four standard face image databases （AR,Yale,ORL and Yale B） indicate that the proposed L-Bagging method is effective and robust to illumination,occlusion and slight pose variation.展开更多
This paper investigates a class of flocks with an M-nearest-neighbor rule,where each agent's neighbors are determined according to M nearest agents with M being a given integer,rather than all the agents within a ...This paper investigates a class of flocks with an M-nearest-neighbor rule,where each agent's neighbors are determined according to M nearest agents with M being a given integer,rather than all the agents within a fixed metric distance as in the well-known Vicsek's model.Such a neighbor rule has been validated by biologists through experiments and the authors will prove that,similar to the Vicsek's model,such a new neighbor rule can also achieve consensus under some conditions imposed only on the system's speed and the number M,n,without resorting to any priori connectivity assumptions on the trajectory of the system.In particular,the authors will prove that if the number M is proportional to the population size n,then for any speed v,the system will achieve consensus with large probability if the population size is large enough.展开更多
文摘为了实现输电塔远距离位移监测,同时满足低成本、无接触、易实施及精确度高等要求,结合输电塔的内轮廓特征和计算机视觉位移识别技术,提出感兴趣区域(region of interest,简称ROI)关键点法。首先,利用N近邻最小能量法进行ROI轮廓搜索提取,并与Harris角点检测算法相结合;其次,通过输电塔台架实验与灰度模板匹配法相比,ROI关键点法位移识别结果的平均误差、均方根误差分别降低了56%和45%,绝对误差小于5 mm和10 mm的准确率提高了61%和3%,计算效率提高了11倍,稳定性及抗噪性能较高;最后,在实验塔对比验证中,ROI关键点法的位移测量值与实际位移的差值百分比在0.0%~11.1%之间。结果表明,ROI关键点法在输电塔结构位移监测中具有较高的准确率、精细度、计算效率、稳定性及鲁棒性。
文摘Bagging is not quite suitable for stable classifiers such as nearest neighbor classifiers due to the lack of diversity and it is difficult to be directly applied to face recognition as well due to the small sample size （SSS） property of face recognition. To solve the two problems,local Bagging （L-Bagging） is proposed to simultaneously make Bagging apply to both nearest neighbor classifiers and face recognition. The major difference between L-Bagging and Bagging is that L-Bagging performs the bootstrap sampling on each local region partitioned from the original face image rather than the whole face image. Since the dimensionality of local region is usually far less than the number of samples and the component classifiers are constructed just in different local regions,L-Bagging deals with SSS problem and generates more diverse component classifiers. Experimental results on four standard face image databases （AR,Yale,ORL and Yale B） indicate that the proposed L-Bagging method is effective and robust to illumination,occlusion and slight pose variation.
基金supported by the National Natural Science Foundation(NNSF)of China under Grant No.61203141the National Center for Mathematics and Interdisciplinary Sciences,Chinese Academy of Science
文摘This paper investigates a class of flocks with an M-nearest-neighbor rule,where each agent's neighbors are determined according to M nearest agents with M being a given integer,rather than all the agents within a fixed metric distance as in the well-known Vicsek's model.Such a neighbor rule has been validated by biologists through experiments and the authors will prove that,similar to the Vicsek's model,such a new neighbor rule can also achieve consensus under some conditions imposed only on the system's speed and the number M,n,without resorting to any priori connectivity assumptions on the trajectory of the system.In particular,the authors will prove that if the number M is proportional to the population size n,then for any speed v,the system will achieve consensus with large probability if the population size is large enough.
