Abstract：
Abstract As one of the short-distance prediction methods for geological conditions in front of the working face, ground penetrating radar has been successfully applied to many tunnel projects at home and abroad. Since no simple and practical software or method for geological interpretation are available, the acquired radar images are often interpreted by experienced engineers. In this paper, a standardized interpretation scheme for radar images is proposed based on field interpretation experience, the relationship between eigenvectors and water content is discovered using a graphic eigenvector extracted with the graphic edge detection operator (Sobel Operator), the feasibility of using an LVQ neural network in the result interpretation of ground penetrating radar is verified by training and simulating with MATLAB software based on its advantages in pattern recognition, and image recognition software is programmed in C# language with good interpretation results achieved in its practical application.
JIN Yong-Rong- Mei-Yuan-Gui
.Application of an LVQ Neural Network to Ground Penetrating Radar Result Interpretations[J] MODERN TUNNELLING TECHNOLOGY, 2013,V50(6): 19-23
2013, Vol. 50 