基于神经网络的隧道地质雷达图像检测技术研究

Research on tunnel geological radar image detection technology based on neural network

利用地质雷达对隧道施工中地质条件及病害进行检测是一种较为常见的手段,但是由于其图像分析工作繁琐且准确率不高,亟待一种智能化检测技术的出现。文章以此为目标,以卷积神经网络为基础构建模型,在引入锚点机制的基础上降低模型对学习数据的需求,并实现相对有效的分析功能。

Using geological radar to detect geological conditions and diseases in tunnel construction is a common means,but due to its cumbersome image analysis and low accuracy,an intelligent detection technology is urgently needed.This paper aims to build a model based on convolutional neural network,reduce the model’s demand for learning data and achieve relatively effective analysis function on the basis of introducing anchor point mechanism.