数字散斑相关方法及其在桥梁检测中的应用研究
文献类型：学位论文
| 作者 | 李凯强; 阎述学
|
| 学位类别 | 硕士 |
| 答辩日期 | 2018-05-17 |
| 授予单位 | 中国科学院沈阳自动化研究所 |
| 授予地点 | 沈阳 |
| 导师 | 朱丹 |
| 关键词 | 数字散斑相关方法 整像素搜索 差分进化 亚像素定位 相机标定 |
| 其他题名 | Research of Digital Speckle Correlation Method and Its Application in Bridge Detection |
| 学位专业 | 控制工程 |
| 中文摘要 | 本文主要围绕数字散斑相关方法中两个核心问题：计算精度和计算速度展开理论研究。首先介绍了目前数字散斑相关方法较为成熟的技术和应用，借助这些先验知识和结论，本文采用高斯光强分布模型实现了计算机生成数字模拟散斑技术；参考现有的散斑质量评价指标，本文综合一阶平均灰度梯度与二阶平均灰度梯度，针对计算机生成的模拟散斑图进行质量评价，实验结果证明了结合两者评价散斑质量的可行性；传统的逐点整像素搜索方法耗费了大量的计算时间，本文实现了基于差分进化算法的整像素搜索方法，并给出了最佳参数设置，提升了计算效率；针对基于九点曲面拟合法的亚像素定位算法，提出了采用六个整像素点的相关系数值直接求解的亚像素定位算法，并与传统曲面拟合法在计算效率和计算精度上进行了比较；基于反向组合型高斯牛顿迭代法的亚像素定位算法采用全局插值系数计算策略耗费计算时间，鉴于此，本文提出了局部区域插值的策略，并在计算速度上同传统迭代法进行了比较。 最后，本文基于提出的改进算法，设计了一套数字散斑相关方法测量系统，借助相机标定方法，针对桥梁检测这一实际应用进行了模拟实验，统计并分析了测量误差。 |
| 英文摘要 | This dissertation mainly focuses on two core issues in the digital speckle correlation method: theoretical calculations and calculation speeds. First, the mature techniques and applications of the digital speckle correlation method are introduced. With the aid of these prior knowledge and conclusions, the Gaussian intensity distribution model is used to implement the computer-generated digital simulation speckle technique; Referring to the existing speckle quality evaluation index, the first-order mean gray gradient and the second-order mean gray gradient are comprehensively considered, evaluating computer-generated simulated speckle pattern quality, and the experimental results demonstrate the feasibility of combining the two indexes to evaluate the quality of speckle images.; The traditional point-by-point integer pixel search method consumes a large amount of computing time. In this paper, an integer-pixel search method based on differential evolution algorithm is implemented, and the optimal parameter setting is given, which improves the computational efficiency; For the sub-pixel localization algorithm based on the nine-point surface fitting method, a sub-pixel localization algorithm is proposed which directly solves the correlation coefficient values of six whole pixels, and compares the calculation efficiency and the calculation accuracy with the traditional surface fitting method.; The sub-pixel localization algorithm based on inverse combined Gaussian Newton iteration method uses the global interpolation coefficient calculation strategy to consume computation time. In view of this, this paper proposes a local area interpolation strategy, and compares it with the traditional iterative method in computation speed. Finally, based on the proposed improved algorithm, a digital speckle correlation measurement system is designed . With the help of the camera calibration method, a simulation experiment is carried out for the actual application of bridge detection, and the measurement error is analyzed and analyzed. |
| 语种 | 中文 |
| 产权排序 | 1 |
| 页码 | 67页 |
| 源URL | [http://ir.sia.cn/handle/173321/21802]  |
| 专题 | 沈阳自动化研究所_水下机器人研究室 |
| 推荐引用方式 GB/T 7714 | 李凯强,阎述学. 数字散斑相关方法及其在桥梁检测中的应用研究[D]. 沈阳. 中国科学院沈阳自动化研究所. 2018. |
入库方式： OAI收割
来源：沈阳自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。