楼道消失点查找与跟踪的迭代算法
提出了一种新的消失点跟踪算法。具体地,改进了梯度霍夫变换,引入映射的参考点;结合梯度方向信息进行映射,提高了映射速度;采用加权最小二乘法在参数空间拟合三角函数曲线,直接计算消失点的坐标。通过迭代的方法更新参考点的坐标,从而使算法精确查找到消失点的位置。实验证明。此算法可以快速准确地查找到楼道环境的消失点,并且迭代速度可以满足消失点的实时跟踪要求。...
Saved in:
| Published in | 计算机应用研究 Vol. 31; no. 3; pp. 735 - 738 |
|---|---|
| Main Author | |
| Format | Journal Article |
| Language | Chinese |
| Published |
State Key Laboratory for High Performance Complex Manufacturing,Changsha 410083,China
2014
College of Mechanical & Electrical Engineering,Central South University,Changsha 410083,China%College of Mechanical & Electrical Engineering,Central South University,Changsha 410083,China |
| Subjects | |
| Online Access | Get full text |
| ISSN | 1001-3695 |
| DOI | 10.3969/j.issn.1001-3695.2014.03.023 |
Cover
| Summary: | 提出了一种新的消失点跟踪算法。具体地,改进了梯度霍夫变换,引入映射的参考点;结合梯度方向信息进行映射,提高了映射速度;采用加权最小二乘法在参数空间拟合三角函数曲线,直接计算消失点的坐标。通过迭代的方法更新参考点的坐标,从而使算法精确查找到消失点的位置。实验证明。此算法可以快速准确地查找到楼道环境的消失点,并且迭代速度可以满足消失点的实时跟踪要求。 |
|---|---|
| Bibliography: | 51-1196/TP corridor vanishing point; gradient Hough transform; Hough transform reference point; least square fitting ZHANG Qiangt, WANG Heng-sheng1'2 ( 1. College of Mechanical & Electrical Engineering, Central South University, Changsha 410083, China; 2. State Key Laboratory for High Per- formance Complex Manufacturing, Changsha 410083, China) This paper proposed a novel algorithm for vanishing point tracking. Specifically, it proposed an improved gradient Hough transform approach, introduced the reference point of mapping,mapped the pixels in image space to the parameter space combined with the gradient direction, which could boost up the Hough transform, and called the weighted least square fitting to fit the trigonometric curve in the parameter space in order todirectly calculate the coordinate of the vanishing point, The algo- rithm could find the vanishing point precisely through iterative method: Experiment results show that this method can find the vanishing point in the corridor fast and accurately and |
| ISSN: | 1001-3695 |
| DOI: | 10.3969/j.issn.1001-3695.2014.03.023 |