跳到主要导航 跳到搜索 跳到主要内容

An improved BP neural network based on GA for 3D laser data repairing

  • Shouqian Yu*
  • , Lixia Rong
  • , Weihai Chen
  • , Xingming Wu
  • *此作品的通讯作者

科研成果: 书/报告/会议事项章节会议稿件同行评审

摘要

Affected by scanning object, environment, scanning speed and user's operation .etc, some information of the object's surface can't be detected by the laser scanner. Aiming at the data loss in laser detecting , the paper presents an improved BP neural network based on GA for 3D laser data repairing, the novelty of this method is adopting Genetic Algorithm(GA) to optimize the configure and weight of network, and at the same time combining Back Propagation(BP) Algorithm to find optimal approximation. The simulation shows the improved BP neural network based on GA has a faster constringency speed and better repairing precision than traditional BP neural network and GA algorithm. Lastly, the paper gives the result of repairing the point cloud collected by 3D information reconstruction system using this network.

源语言英语
主期刊名2008 IEEE International Conference on Robotics, Automation and Mechatronics, RAM 2008
571-576
页数6
DOI
出版状态已出版 - 2008
活动2008 IEEE International Conference on Robotics, Automation and Mechatronics, RAM 2008 - Chengdu, 中国
期限: 21 9月 200824 9月 2008

出版系列

姓名2008 IEEE International Conference on Robotics, Automation and Mechatronics, RAM 2008

会议

会议2008 IEEE International Conference on Robotics, Automation and Mechatronics, RAM 2008
国家/地区中国
Chengdu
时期21/09/0824/09/08

指纹

探究 'An improved BP neural network based on GA for 3D laser data repairing' 的科研主题。它们共同构成独一无二的指纹。

引用此