Estimating the Propagation Velocity Distribution and the Position of Buried Pipe Using x -t - v Data Matrix of Subsurface Radar Image

We propose a new method of radar imaging of underground pipes in inhomogeneous soil, which does not require a priori knowledge of the propagation velocity of electromagnetic waves in soil. In this method, three-dimensional x t -v data matrix from multiple migrated radar images with different propagation velocities using F-K migration method is composed. And we extract the hyperbolic apex and estimate the propagation velocity of each hyperbola, using the x t -v data matrix. By interpolating these estimated discrete propagation velocities along the t axis, the propagation velocity distribution downward in the soil is derived. The reconstructed image of the underground section in inhomogeneous soil is obtained by extracting the image intensity from the x t -v data matrix based on the derived propagation velocity distribution. Furthermore the influences by thick buried pipes are considered, and the diameter estimation method is described. The effectiveness of new method is demonstrated by comparison with previous methods. The new method accurately reconstructs the underground section in inhomogeneous soil a t low computing cost.