| 基于目标跟踪及神经优化的运动特征点对应 |
| |
| 引用本文： | 汪亚明. 基于目标跟踪及神经优化的运动特征点对应[J]. 中国机械工程, 2001, 12(11): 1264-1266 |
| |
| 作者姓名： | 汪亚明 |
| |
| 作者单位： | 浙江工程学院信息电子学院 (杭州市 310033) |
| |
| 基金项目： | 浙江省教委科研项目 ( 2 0 0 0 0 36) |
| |
| 摘 要： | 将目标跟踪及神经优化技术应用特征点对应之中。根据图像序列的特点，用目标跟踪技术得到特征点的顶测位置，然后通过Hopfield网络最小化特征点间的费用函数使帧间特征点取得正确的对应，由于本方法没有涉及任何刚体运动约束，所以适用于各种非刚体运动的特征点对应。
|
| 关 键 词： | 机器视觉 图像序列 特征点对应 目标跟踪 Hopfield网络 神经优化 |
Motion Feature Point Correspondence Based on Target Tracking and Neural Optimization |
| |
| WANG Yaming. Motion Feature Point Correspondence Based on Target Tracking and Neural Optimization[J]. China Mechanical Engineering, 2001, 12(11): 1264-1266 |
| |
| Authors: | WANG Yaming |
| |
| Abstract: | This paper proposes a new correspondence method based on the target tracking and neural network. According to the characteristics of the image sequence, the predicted position can be obtained through the technique of target tracking and the correct correspondence can be established through minimizing the cost function between frames using Hopfield neural network. Because this method is not based on the constraint of body rigidity, it can solve the correspondence problem of nonrigid body motion. |
| |
| Keywords: | machine vision image sequence feature point correspondence target tracking Hopfield neural network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
