The accurate estimation of GPR migration velocity and comparison of imaging methods

In order to improve the resolution of GPR profiles from the perspective of data processing, the migration technique is applied to forward and measured GPR profiles. Given the shortcomings of the conventional GPR migration velocity determination method, such as strong randomness, uncontrollable error and massive consumption of workforce and material resources, therefore, a new method for determining GPR migration velocity based on cross-correlation analysis and least square fitting has been proposed. According to the determined velocity, three migration algorithms (Kirchhoff, FK and finite difference) can be used to process the forward and measured GPR profiles (pipes detection), the migration effect evaluated from image display, the information entropy, the operational efficiency and the accuracy of the determined velocity also reflected from migration results. At the same time, the radius inversion is realized under the coordinate points picked up by the cross-correlation analysis and fitted equations. In addition, a series of studies show that the migration velocity determined by the method proposed in this paper can be used not only to migrate GPR profiles, but to converge the hyperbola in profiles completely, which indicates the velocity is relatively accurate. By analyzing the effect of the GPR profiles migrated by three algorithms, it can be found that the FK algorithm has a strong applicability to the forward and measured profiles of GPR and that the radius inversion errors of forward and measured profiles can be controlled below 6% and 12% respectively.