System and method for automatic quality control for assembly line processes
AssigneeTaiwan Semiconductor Manufacturing Company, Ltd.
InventorsKewei ZUO, Chien Rhone WANG, Tzu-Cheng LIN, Chih-Wei LAI
Life of this patent
Estimating the term…
Abstract
A system and method is disclosed for a quality control and/or inspection procedure for assembly line processes. The disclosed system and method enable automatic optical inspection of a device during different stages of manufacture as well as in its finished form. The disclosed system and method enable the automatic quality control process to be self-learning, dynamic, and to identify and classify defects in real time.
Claims20
A method for an inspection procedure, the method comprising the steps of:
- (a) providing a device having a plurality of features;
- (b) obtaining a first image of the device;
- (c) providing a second image wherein the second image is a benchmark image for the device;
- (d) characterizing each of the plurality of features in the first image using a first predetermined criteria;
- (e) characterizing each of a plurality of features in the second image using the first predetermined criteria;
- (f) comparing the characterization of one of the plurality of features in the first image with the characterization of the corresponding feature in the second image;
- and
- (g) creating an output image comprising the first image with a highlighted area wherein the highlighted area comprises a defect determined from the comparison.
A method for an inspection procedure, the method comprising the steps of:
- (a) providing a device having a plurality of features;
- (b) obtaining a first image of the device;
- (c) processing the first image using a predetermined background filter, wherein the background filter is selected from the group consisting of:
- a binary filter, a gray scale filter, an image enhancement filter, and combinations thereof;
- (d) performing a normalizing procedure of the first image, wherein the normalizing procedure is selected from the group consisting of:
- image resolution processing, scale processing, device rotation processing, and device tilt angle processing;
- (e) providing a second image wherein the second image is a benchmark image for the device;
- (f) characterizing each of the plurality of features in the first image using a first predetermined criteria;
- (g) characterizing each of a plurality of features in the second image using the first predetermined criteria;
- (h) comparing the characterization of one of the plurality of features in the first image with the characterization of the corresponding feature in the second image;
- (i) creating an output image comprising the first image with a highlighted area wherein the highlighted area comprises a defect determined from the comparison;
- (j) repeating step (h) for each of the plurality of features;
- (k) repeating steps (f) through (h) and step (j) using a second predetermined criteria;
- (l) classifying each defect by a type of defect according to a predetermined classification system;
- (m) creating a report including a number of defects by type of defect;
- (n) correlating a first data in the report with a second data from a second inspection method;
- and
- (o) updating a knowledge database with at least one of a result of the comparison of characterization of the one feature and a result of the correlation.
A system for inspecting a device having a plurality of features, the system comprising:
- a C-mode scanning acoustic microscope (“C-SAM”) to obtain a first image of said device;
- a processor operably connected to said C-SAM wherein said processor is programmed to:
- characterize each of the plurality of features in said first image using a first predetermined criteria;
- characterize each of a plurality of features in a second image of said device using the first predetermined criteria, wherein said second image is a benchmark image for said device;
- compare the characterization of one of the plurality of features in said first image with the characterization of the corresponding feature in said second image to thereby identify a defect;
- classify said defect by a type of defect according to a predetermined classification system;
- and
- create a report including a number of defects by type of defect;
- a database operably connected to said processor wherein said database stores said benchmark image and an updated version of the benchmark image;
- and
- imaging circuitry operably connected to said processor for creating an output image comprising the first image with a highlighted area wherein the highlighted area comprises the defect.
Dependencies are inferred from the claim text.