Iterative Detection of Linear Objects in GPR and Seismic Images

An iterative technique for detecting linear features in 2D images based on the Radon transform is developed. The algorithm is suitable for processing ground penetrating radar (GPR) and seismic images to find underground pipes and tunnels. Detection is performed in the Radon transform space. The length and width of the linear object area are estimated and the corresponding portion of the image is removed at each iteration. The algorithm has the advantage of detecting very weak linear objects, which are barely detected by existing detection algorithms, in Radon transform space. The proposed algorithm is tested on both simulated and experimental data measured from a laboratory scale model area. Results show successful detection of strong and weak linear objects