作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2009, Vol. 35 ›› Issue (8): 205-207. doi: 10.3969/j.issn.1000-3428.2009.08.069

• 人工智能及识别技术 • 上一篇    下一篇

基于遗传算法的直线光流刚体运动重建

杨建伟,陈 震,危水根,简宏伟   

  1. (南昌航空大学自动化学院计算机视觉研究室,南昌 330063)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2009-04-20 发布日期:2009-04-20

Recover 3-D Motion of Rigid Body from Straight-line Optical Flow Based on Genetic Algorithm

YANG Jian-wei, CHEN Zhen, WEI Shui-gen, JIAN Hong-wei   

  1. (Laboratory of Computer Vision, College of Automation, Nanchang Hangkong University, Nanchang 330063)
  • Received:1900-01-01 Revised:1900-01-01 Online:2009-04-20 Published:2009-04-20

摘要: 建立一种新的基于直线光流场从单目图像序列恢复刚体运动和结构的模型,推导出直线光流场与刚体的运动参数之间的关系,用2个二阶线性微分方程表达这种关系,并提出一种求解刚体运动参数的遗传算法,只需要获得图像平面的2条直线光流即可求解刚体的旋转参数,并用合成图像测试了该算法的有效性。

关键词: 单目图像序列, 直线光流, 遗传算法

Abstract: This paper presents a new model for recovering the 3-D structure and motion using the image sequences of monocular images. It deductes the relationship between straight line optical flow and the motion of a rigid body, and represents it with second derivative linear equation. A new method is presented to solve the model based on Genetic Algorithm(GA). It needs at least two group of straight-line optical flow in projection plane to recover rotational components of motion. Validity of the algorithm is tested with synthetic sequences.


Key words: sequences of monocular images, straight-line optical flow, Genetic Algorithm(GA)

中图分类号: