Processing and Analysis on the GPR Image of Dam Hidden Hazard by Means of Artificial Neural Network

摘要

The GPR effect of detecting dam hidden hazard relies on recognition of hidden hazard information in reflection image.The computing flow is introduced that processing GPR image by means of self-fitting ANN on MATLAB platform.Then the comparison results of practical GPR images between before and after processing using ANN are presented.The application results demonstrate that the hidden hazard information can be more obvious and the interpretation result can be more accurate through ANN method.And ANN method is able to improve the resolution of GPR image.The analysis result of practical image for one leak detection project is identical to that of the on-site excavation.And the grouting quantity for the interpreted non-uniform zone is more than that in other zone.So it welt helped the seepage treatment proiect.