Systems and methods for identifying waveguide defects
Systems and methods are described herein for waveguide defect detection. In some examples, a statistical value of pixels of a number of image locations in one or more images of the waveguide can be computed. An extremity matrix can be generated based on an anomaly detector model and statistical pixel data that includes statistical values computed for respective image locations in the one or more images of the waveguide. A decision can be made to determine whether the waveguide is a conforming or a non-conforming waveguide based on the extremity matrix and a set of thresholds. In some examples, the statistical value is a mean intensity value or an average intensity value.
1 . A computer-implemented method carried out with one or more processors comprising:
receiving image data from an infra-red (IR) sensor in response to injecting a radio frequency test signal into a waveguide under test, the image data comprising one or more images of the waveguide under test;
computing, using a processor, a statistical value of pixels of a number of image locations in the one or more images of the waveguide;
generating, using a processor, an extremity matrix based on an anomaly detector model and statistical pixel data that includes statistical values computed for respective image locations in the one or more images of the waveguide; and
determining, using a processor, whether the waveguide is a conforming or a non-conforming waveguide based on the extremity matrix and a set of thresholds.
2 . The computer-implemented method of claim 1 , wherein the anomaly detector model comprises a number of probability distributions, each probability distribution being computed based on a respective statistical set of statistical sets, wherein each statistical set of the statistical sets identifies a respective image location of a number of image locations across one or more images of a number of conforming waveguides, and a statistical value for the respective image location.
3 . The computer-implemented method of claim 2 , further comprising fitting a distribution to each of the statistical sets to provide one of the probability distributions of the anomaly detector model.
4 . The computer-implemented method of claim 3 , wherein the distribution is a continuous probability distribution.
5 . The computer-implemented method of claim 2 , wherein the extremity matrix includes a number of deviation values, each deviation value corresponding to a deviation of one of the statistical values for the respective image locations in the one or more images of the waveguide from a median or an average of a corresponding probability distribution of the probability distributions of the anomaly detector model.
6 . The computer-implemented method of claim 5 , wherein determining whether the waveguide is a conforming or a non-conforming waveguide comprises evaluating each deviation value of the extremity matrix relative to the set of thresholds.
7 . The computer-implemented method of claim 6 , further comprising outputting anomaly data indicating that the waveguide is a conforming or non-conforming waveguide based on the evaluation.
8 . The computer-implemented method of claim 7 , wherein the anomaly data identifies one or more image locations in the one or more images of the waveguide corresponding to a potential image anomaly location, the potential image anomaly location being indicative of a potential defect at a respective location on the waveguide.
9 . The computer-implemented method of claim 8 , further comprising:
receiving user data indicating whether a defect is present at the respective location; and
updating the anomaly detector model based on the user data.
10 . The computer-implemented method of claim 9 , wherein updating the anomaly detector model based on the user data comprises adjusting one or more parameters of at least one probability distribution of the probability distributions of the anomaly detector model.
11 . The computer-implemented method of claim 1 , wherein the one more images of the waveguide are one or more infrared (IR) images, the method further comprising processing the image data to convert color pixels of the one or more IR images of the image data to a matrix of radio-frequency (RF) amplitude values, the statistical value of the pixels being computed based on the matrix of RF amplitude values.
12 . A system comprising:
memory to store machine-readable instructions and data comprising an anomaly detector model comprising pre-computed probability distributions;
one or more processors to access the memory and execute the machine-readable instructions, the machine-readable instructions comprising:
a statistical value calculator programmed to compute a statistical value of pixels of each image location in one or more images of a waveguide under test, wherein the one or more images are received in response to injection of one or more radio frequency (RF) signals into the waveguide under test;
a deviation calculator programmed to generating an extremity matrix based on an evaluation of the statistical value of the pixels of each image location in the one or more images of the waveguide and a respective one of the pre-computed probability distributions, wherein the extremity matrix includes a number of deviation values; and
a decision function programmed to determine whether the waveguide is a conforming or a non-conforming waveguide based on the deviation values of the extremity matrix and relative to a set of thresholds.
13 . The system of claim 12 , each deviation value corresponding to a deviation of one of statistical values for respective image locations in the one or more images of the waveguide from a median of a corresponding pre-computed probability distribution of the pre-computed probability distributions of the anomaly detector model.
14 . The system of claim 13 , wherein the decision function is further programmed evaluate each deviation value of the extremity matrix relative to the set of thresholds and output anomaly data indicating that the waveguide is a conforming or non-conforming waveguide based on the evaluation of deviation values of the extremity matrix relative to the set of thresholds.
15 . The system of claim 14 , wherein the anomaly data identifies one or more image locations in the one or more images of the waveguide corresponding to a potential image anomaly location, the potential image anomaly location being indicative of a potential defect at a respective location on the waveguide, the decision function being further programmed to receive user data indicating whether a defect is present at the respective location, and update the anomaly detector model based on the user data.
16 . The system of claim 15 , wherein updating the anomaly detector model based on the user data comprises adjusting one or more parameters of at least one pre-computed probability distribution of the pre-computed probability distributions of the anomaly detector model.
17 . The system of claim 16 , wherein the one more images of the waveguide are one or more infrared (IR) images and the waveguide is a radio-frequency (RF) waveguide.
18 . A system comprising:
a waveguide under test;
an infra-red (IR) sensor for generating image data, the image data comprising one or more images of the waveguide under test;
a test driver for injecting the waveguide under test with one or more radio frequency signals;
memory to store machine-readable instructions and data comprising an anomaly detector model comprising pre-computed probability distributions;
one or more processors to access the memory and execute the machine-readable instructions, the machine-readable instructions comprising:
a statistical value calculator programmed to compute a statistical value of pixels of each image location in the one or more images of the waveguide under test, wherein the one or more images are received by the IR sensor in response to injection of the one or more RF signals into the waveguide under test;
a deviation calculator programmed to generating an extremity matrix based on an evaluation of the statistical value of the pixels of each image location in the one or more images of the waveguide and a respective one of the pre-computed probability distributions, wherein the extremity matrix includes a number of deviation values; and
a decision function programmed to determine whether the waveguide is a conforming or a non-conforming waveguide based on the deviation values of the extremity matrix relative to a set of thresholds.
19 . The system of claim 18 , wherein the one more images of the waveguide are one or more infrared (IR) images, the method further comprising processing the image data to convert color pixels of the one or more IR images of the image data to a matrix of radio-frequency (RF) amplitude values, the statistical value of the pixels being computed based on the matrix of RF amplitude values.
20 . The system of claim 18 , wherein each deviation value corresponding to a deviation of one of statistical values for respective image locations in the one or more images of the waveguide under test from a median of a corresponding pre-computed probability distribution of the pre-computed probability distributions of the anomaly detector model.