System and method for automated borescope inspection user interface

1. A method of improving automated defect detection, the method comprising:

providing an image capture device for capturing and transmitting video images of an object in motion in a device;

performing automated Robust Principal Component Analysis on the video images to simultaneously decompose the video images into a low rank matrix representing a normal object and a sparse matrix representing an object anomaly;

processing the sparse matrix to determine potential defects or potential non-defects;

reviewing results of the automated analysis by a human inspector and providing feedback; and

refining the automated analysis on the data based upon the feedback of the human inspector.

2. The method of claim 1 , wherein performing the automated analysis of the data comprises:

classifying the potential defects or potential non-defects into defects or non-defects.

3. The method of claim 2 , wherein classifying the potential defects or the potential non-defects comprises:

providing a classifier that utilizes the potential defects or the potential non-defects from the Robust Principal Component Analysis as an input; and

generating at least one of a certainty, an uncertainty and a binary output to classify the potential defects or the potential non-defects into the defects or the non-defects.

4. The method of claim 3 , wherein the classifier outputs a binary value of one if the classifier classifies the potential defects or the potential non-defects into the defects.

5. The method of claim 3 , wherein the classifier outputs a binary value of zero if the classifier classifies the potential defects or the potential non-defects into the non-defects.

6. The method of claim 1 , wherein the object is a plurality of jet engine blades within one stage of the jet engine.

7. The method of claim 1 , wherein reviewing the results of the automated analysis by a human inspector comprises one of accepting and rejecting the results of the automated analysis.

8. The method of claim 7 , wherein accepting the results of the automated analysis comprises saving the results within a confirmed defect database.

9. The method of claim 7 , wherein rejecting the results of the automated analysis comprises saving the results within a dismissed defect database.

10. The method of claim 1 , wherein reviewing the results of the automated analysis by a human inspector comprises indicating further manual or automated analysis to be performed on the data.

11. The method of claim 1 , wherein refining the automated analysis comprises updating one of a confirmed defect database and a dismissed defect database and revising the automated analysis based on the updated database to improve accuracy of the automated analysis.

12. A system for performing automated defect detection, the system comprising:

an image capture device configured to capture and transmit video images of one or more moving components of an object; and

a monitoring and analysis site in at least indirect communication with the image capture device, the monitoring and analysis site configured to perform an automated analysis of the video images using Robust Principal Component Analysis on the video images to simultaneously decompose the video images into a low rank matrix representing a normal component and a sparse matrix representing a component anomaly and to process the sparse matrix to determine potential defects or potential non-defects; and

a database configured to store results of the automated analysis.

13. The system of claim 12 , wherein the object is at least one of an engine and a turbine comprising a plurality of stages, each of the plurality of stages having a plurality of blades.

14. The system of claim 12 , wherein the monitoring and analysis site is a remote site.

15. A method of performing automated defect detection, the method comprising:

providing an image capture device configured to capture and transmit a sequence of video images of one or more rotating blades of an engine;

performing an automated analysis on the sequence of video images, the automated analysis comprising (a) performing a Robust Principal Component Analysis on the sequence of video images to simultaneously decompose the video images into a low rank matrix representing a normal blade and a sparse matrix representing a blade anomaly and to generate analysis results by processing the sparse matrix to determine potential defects or potential non-defects; and (b) utilizing a classifier to classify the analysis results into defects or non-defects by generating an output;

verifying the defects or the non-defects by a human inspector and providing feedback; and

re-training the classifier based upon the feedback from the human inspector.

16. The method of claim 15 , wherein the defects comprises one or more of a type of the defect, a location of the defect, and a size of the defect.

17. The method of claim 16 , wherein the type of the defect may be one or more of leading edge defects, erosions, nicks, cracks, dents, and cuts.

18. The method of claim 15 , wherein providing feedback comprises one of (a) accepting the output of the classifier and storing the output into a confirmed defect database; (b) rejecting the output of the classifier and storing the output into a dismissed defect database; and (c) indicating further manual or automated analysis to be performed.