基于SIFT图像匹配算法的桥梁变形室内试验研究

Laboratory experimental study on bridge deformation based on SIFT image matching algorithm

全景拍摄图像处理在桥梁变形监测中受透视效应影响较大,故提出了旋转拍摄拼接方法。为研究其精度,制备了钢筋混凝土梁模型,通过四点弯曲试验模拟桥梁变形状态。数据采集方式包括百分表测量、全景拍摄和旋转拍摄。通过SIFT算法实现相邻图像匹配拼接工作,对两种方法测量结果进行对比分析。室内试验研究结果表明,旋转拍摄在变形测量上具有更高的精度,利用SIFT函数进行特征点提取及匹配,可以得到连续图像的高精度拼接效果图。本研究为旋转图像拍摄方法实时监测桥梁变形提供了实验依据。

Panoramic shooting image processing is greatly affected by perspective effect in bridge deformation monitoring, so a rotational shooting stitching method is proposed. In order to study its accuracy, a reinforced concrete beam model was prepared, and the deformation state of the bridge was simulated by a four-point bending test. Data acquisition methods include dial gauge measurement, panoramic shooting and rotation shooting. The adjacent image matching and stitching work is realized by the SIFT algorithm, and the measurement results of the two methods are compared and analyzed. The results of indoor experiments show that rotational shooting has higher accuracy in deformation measurement, and the SIFT function is used to extract and match feature points, and high-precision stitching effect maps of continuous images can be obtained. This study provides an experimental basis for the rotational image shooting method to monitor the deformation of the bridge in real time.