基于视觉数据融合和机器学习算法的在役桥梁病害智能检测方法

为了准确检测在役桥梁病害问题,研究了基于视觉数据融合和机器学习算法的在役桥梁病害智能检测方法。采用雷达与视觉数据融合的无人机航拍技术,采集和定位在役桥梁高清图像;构建基于Ｍask R-CNN卷积神经网络的在役桥梁病害检测模型,通过迁移学习法动态优化训练检测模型,得出最优的在役桥梁病害检测模型;对定位到的在役桥梁高清图像进行去噪增强处理,输入该检测模型进行分类识别,输出病害检测结果。实验结果表明：该方法可以快速准确检测出多种在役桥梁病害,减少误检和漏检,实现一图多病害的精细化检测。

Intelligent Detection Method of Bridge Diseases in Service Based on Visual Data Fusion and Machine Learning Algorithm

In order to accurately detect the existing bridge diseases, an intelligent detection method of existing bridge diseases based on visual data fusion and machine learning algorithm is studied. UAV aerial photography technology of radar and visual data fusion is adopted to collect and locate the high-definition image of the bridge in service.An in-service bridge disease detection model based on Mask R-CNN convolution neural network is built to train the detection model dynamically through the migration learning method to obtain the optimal in-service bridge disease detection model.To de-noise and enhance the located high-definition image of the bridge in service, input the detection model for classification and recognition, and output the disease detection results. The experimental results show that this method can quickly and accurately detect a variety of bridge diseases in service, reduce false detection and missed detection, and realize the fine detection of multiple diseases in one map.