欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对双目视觉中两摄像机的固定约束关系,提出了基于双目视觉的光束法平差新算法。对传统光束法平差法方程构造过程进行了改进,使每次迭代过程中只需对拍摄的两幅图像中一个摄像机的外参数进行优化,更好地将两个摄像机绑定为一个摄像机,使优化精度更高;新算法中,法化矩阵的维数比传统算法降低了12,所以每次迭代计算所需要的时间也会减少,对于迭代次数越多的优化过程,新算法和传统算法计算所需时间差越明显。模拟实验和实际测量均表明,新算法比传统算法优化精度高,计算所需时间短。
英文摘要：
In this paper,we propose a new algorithm for bundle adjustment based on the fixed constraints between the two cameras in stereo vision.The new algorithm improves the normal equation of traditional bundle adjustment,and it only needs to optimize extrinsic parameters of one camera in each iteration.The two cameras can be better bounded as one and it can achieve a higher accuracy.The dimension of the matrix in the new algorithm is half of that in the traditional algorithm.Therefore it will reduce the time of calculation in each iteration.Simulation and actual measurement experiments both prove that the accuracy of the new algorithm is higher than that of the traditional algorithm and the new algorithm costs shorter time.
同期刊论文项目
同项目期刊论文
期刊信息
位置：