Automated depth estimation for non-destructive testing
A method for estimating a depth of one or more buried objects can include receiving, from a non-destructive testing (NDT) scanner, a B-Scan image of a scanned region. Using a machine learning model, a hyperbola in the B-Scan image can be identified. At least one geometric feature of the hyperbola can be determined. Based upon the at least one geometric feature of the hyperbola, a dielectric constant in at least a portion of the scanned region can be determined. A depth of the one or more objects beneath a surface of the scanned region can be determined based upon the determined dielectric constant.
1 . A method comprising:
receiving, from a non-destructive testing (NDT) scanner, a B-Scan image of a scanned region;
identifying, using a machine learning model, a set of pixels associated with a hyperbola in the B-scan image;
determining at least one geometric feature of the set of pixel;
determining, based upon the at least one geometric feature of the set of pixels, a dielectric constant in at least a portion of the scanned region; and
determining, based upon the dielectric constant, a depth of one or more objects beneath a surface of the scanned region.
2 . The method of claim 1 , wherein the machine learning model is a computer vision model.
3 . The method of claim 2 , wherein the machine learning model is an instance segmentation model.
4 . The method of claim 1 , wherein determining the at least one geometric feature of the set of pixels comprises:
determining a center line of the set of pixels;
determining a function that represents the center line.
5 . The method of claim 1 , wherein the at least one geometric feature comprises an apex location value, a curvature value, an amplitude value, or a combination thereof.
6 . The method of claim 1 , further comprising training the machine learning model using training data comprising a plurality of B-Scan images from a geographic location in which the scanned region is located.
7 . The method of claim 1 , further comprising training the machine learning model using training data comprising a plurality of B-Scan images generated by NDT scanners similar to the NDT scanner from which the B-Scan image was received.
8 . The method of claim 1 , further comprising, prior to identifying the set of pixels using the machine learning model, processing the B-Scan image.
9 . The method of claim 8 , wherein processing the B-Scan image comprises reducing noise in the B-Scan image, enhancing a contrast of the B-Scan image, or a combination thereof.
10 . The method of claim 8 , wherein processing the B-Scan image comprises:
receiving, from an inertial measurement unit (IMU) coupled to the NDT scanner, position data associated with the NDT scanner; and normalizing the B-Scan image based upon the position data.
11 . The method of claim 1 , further comprising providing information indicating the dielectric constant and the depth of the object to a user.
12 . The method of claim 11 , wherein providing the information indicating the dielectric constant and the depth of the object to a user comprises:
wirelessly transmitting the information indicating the dielectric constant and the depth of the object to a processor of the NDT scanner; and
displaying the dielectric constant and the depth of the object to the user by a user interface of the NDT scanner.
13 . A system comprising:
at least one processor; and
non-transitory memory storing instructions that, when executed by the at least one processor, cause the at least one processor to perform operations comprising:
receiving, from a non-destructive testing (NDT) scanner, a B-Scan image of a scanned region;
identifying, using a machine learning model, a set of pixels associated with a hyperbola in the B-scan image;
determining at least one geometric feature of the set of pixels;
determining, based upon the at least one geometric feature of the set of pixels, a dielectric constant in at least a portion of the scanned region; and
determining, based upon the dielectric constant, a depth of an object beneath a surface of the scanned region.
14 . The system of claim 13 , wherein the machine learning model is a computer vision model.
15 . The system of claim 14 , wherein the machine learning model is an instance segmentation model.
16 . The system of claim 13 , wherein determining the at least one geometric feature of the set of pixels comprises:
determining a center line of the set of pixels; and
determining a function that represents the center line.
17 . The system of claim 13 , wherein the at least one geometric feature comprises an apex location value, a curvature value, an amplitude value, or a combination thereof.
18 . The system of claim 13 , further comprising the NDT scanner.
19 . The system of claim 13 , wherein the processor and the non-transitory memory are components of the NDT scanner.
20 . A non-transitory computer readable storage medium storing instructions that, when executed by at least one processor, cause the at least one processor to perform operations comprising:
receiving, from a non-destructive testing (NDT) scanner, a B-Scan image of a scanned region;
identifying, using a machine learning model, a set of pixels associated with a hyperbola in the B-scan image;
determining at least one geometric feature of the set of pixels;
determining, based upon the at least one geometric feature of the set of pixels, a dielectric constant in at least a portion of the scanned region; and
determining, based upon the dielectric constant, a depth of an object beneath a surface of the scanned region.
21 . A method comprising:
receiving, from a non-destructive testing (NDT) scanner, a B-Scan image of a scanned region;
identifying, using a machine learning model, a set of pixels associated with a hyperbola in the B-scan image, wherein the machine learning model is an instance segmentation model;
determining at least one geometric feature of the set of pixels, wherein the at least one geometric feature comprises an apex location value;
determining, based upon the at least one geometric feature of the set of pixels, a dielectric constant in at least a portion of the scanned region; and
determining, based upon the dielectric constant, a depth of one or more objects beneath a surface of the scanned region;
wirelessly transmitting the information indicating the dielectric constant and the depth of the object to a processor of the NDT scanner; and
displaying the dielectric constant and the depth of the object to the user by a user interface of the NDT scanner.