基于多目视觉的变光照环境无靶标拉索三维振动测量原理与方法

动态位移和振动频率是反映拉索服役状态的关键指标。常用测试手段存在测点少、三维位移测量困难等局限，无法精确表征现役拉索的振动特性。本项目拟建立一套以多目视觉自标定和曲线特征映射重构为基础的无靶标拉索动态位移测量理论及振动特性计算方法，实现现场变光照环境下多根拉索全场三维振动的同步测量。主要研究内容包括：①发展无线动态同步采集技术，研制多目测量系统；建立基于共同场景约束的系统全参数自标定方法，探究标定过程中误匹配点剔除及全局优化算法。②提出空间—平面曲线特征映射重构方法，探明以全时域场景为基础的无靶标拉索三维动态位移测量原理，并利用频谱分析计算其振动特性。③明确变光照环境对视觉测量的影响规律，研究基于曝光优化控制和图像增强的光照改善机制，从而提升测量系统的光照鲁棒性。本项目预期解决现役拉索全场三维振动测量的难题，有助于拉索风致振动和疲劳失效机理研究，对提升索结构桥梁可靠性和耐久性具有重要意义。

Dynamic displacement and vibration frequency are among the key indicators that represent the service status of the cables. Due to the less points of interests and the difficulties in three-dimensional (3D) displacement measurement, the current testing methods cannot accurately characterize the vibration of cables in service. In order to obtain the full-field 3D vibration of multiple cables simultaneously under illumination varying condition, this project intends to establish the dynamic displacement measurement theory and vibration characteristics calculation method for target-free cables, which is based on self-calibration of multi-camera vision system and curve feature mapping reconstruction. The main research contents are as follows: (1) The wireless-based dynamic synchronous acquisition technology, which can capture images simultaneously for different camera, is developed for multi-camera vision system. Then, some geometries in the common scene are regarded as the constraints for internal and external parameters estimation during the self-calibration process. Besides, the mismatching point eliminating and the combinatorial optimization algorithms are investigated for obtaining global optimized parameters. (2) The principle of 3D reconstruction, namely, the space-to-plane curve feature mapping method is presented and studied. After that, time-sequence images and spectrum analyses are adopted for calculating the dynamic displacements and vibration characteristics of target-free cables. (3) The influence of varying illumination on visual system is clarified. For enhancing the illumination robustness of multi-camera vision system, some optimization strategies are proposed based on auto exposure control and image enhancement. In summary, this project is expected to solve the problems of 3D vibration measurement of the target-free cables in service, which is beneficial to the mechanism research of the wind-induced vibration and fatigue failure. It is also of great significance for the improvement of the reliability and durability of the cable-supported bridges.