Detecting a shallowly buried and elongated object estimating its orientation using commonly adopted co-polarized GPR system is challenging due to the presence of strong ground clutter that masks target reflection. A cross-polarized configuration can be used suppress reveal reflection, but it suffers from inconsistent detection capability which significantly varies with different orientations. T...

Using a fully vectorial three-dimensional numerical approach (generalized field propagator, based on Green’s tensor technique), we investigate the near-field images produced by subwavelength objects buried in a dielectric surface. We study the influence of the object index, size, and depth on the near field. We emphasize the similarity between the near field spawned by an object buried in the s...

Among the ultrasound based NDT methods and techniques, some are used for materials characterization. In a typical situation, some of the measurable properties of a tested object’s material are used as a data set for characterization. The basis of measurements are properties of an ultrasonic pulse that can propagate through the object. In order to obtain intense enough registered ultrasonic puls...

Motivated by a 3D printing company, ExOne, the goal of this project is to recognize 3D objects from partially observed surface data. Given point cloud data of a buried object from a Kinect, as well as the entire CAD model of the object, the aim is to determine the object's orientation in space. To accomplish this task, we implemented a Hough Voting algorithm for object detection and localizatio...

The impulse radar has shown promising results for detecting buried objects, even non-metallic ones. One problem with ground penetrating radars is the strong backscattered signal from the ground surface. If the object is buried deeply below the surface this is no problem since the backscattered signal from the surface will arrive earlier than the target signal and we only have to gate the time s...

Many effective supervised discriminative dictionary learning methods have been developed in the literature. However, when training these algorithms, precise ground-truth of the training data is required to provide very accurate point-wise labels. Yet, in many applications, accurate labels are not always feasible. This is especially true in the case of buried object detection in which the size o...

GPR is a nondestructive method to scan the subsurface. On the resulting radargrams, originally interpreted manually in a time consuming process, one can see hyperbolas corresponding to buried objects. For accelerating the interpretation a machine shall be enabled to recognize hyperbolas on radargrams autonomously. One possibility is the combination of clustering with an expectation maximization...

Keeping track of underground utilities through maps or real physical signs is essential for their maintenance and quick repairs, whenever required, without causing much obstruction to day to day life. It is not uncommon that maps are misplaced or real physical signs are destroyed. In such situations, digging and excavation becomes unavoidable during repair works. Ground penetrating radar (GPR) ...