Manufacturing Quality Improvement Through Statistical Root Cause Analysis Using Convolution Neural Networks
AssigneeDpix, LLC
InventorsJerome David Crocco, Paul R. O'Hern, JR.
Life of this patent
Estimating the term…
Abstract
A quality improvement method for an image sensor array includes collecting in-line optical inspection data of the image sensor, collecting end of line electrical data of the image sensor, creating defect maps and obtaining x-y coordinates of the optical inspection data and the electrical data, correlating the defect maps to generate correlated defects, classifying the images of the correlated defects, and generating root cause statistics of the classified correlated defects.
Claims20
A classification system of root cause failure in an image sensor array comprising:
- an optical inspection system for generating inspection data of the image sensor;
- a tester for generating test data of the image sensor;
- and
- a computer system configured for:
- generating two dimensional coordinates of optical defects present in the inspection data;
- generating two dimensional coordinates of electrical failures in the test data;
- correlating defects between the optical defects and the electrical failures;
- classifying the correlated defects;
- and
- providing root cause data based on the classified correlated defects.
An in-line prediction method for an image array of sensors comprising:
- collecting in-line optical inspection data of the image array of sensors;
- passing the optical inspection data through an image classifier to generate a plurality of defect types;
- selecting defect types that most strongly affect manufacturing yield of the image sensor;
- and
- generating a pass/fail prediction based on the selected defect types.
A quality improvement method for an image sensor comprising:
- collecting in-line optical inspection data of the image sensor;
- collecting end of line electrical data of the image sensor;
- creating defect maps and obtaining x-y coordinates of the optical inspection data and the electrical data;
- correlating the defect maps to generate correlated defects;
- classifying the images of the correlated defects;
- and
- generating root cause statistics of the classified correlated defects.
Dependencies are inferred from the claim text.