Method and apparatus for rapidly classifying defects in subcomponents of manufactured component
The present disclosure provides methods and apparatus for rapidly classifying detected defects in subcomponents of a manufactured component or device. The defect classification may occur after defect detection or, because the classification may be sufficiently rapid to be performed in real-time, during defect detection, as part of the defect detection process. In an exemplary implementation, the presently-disclosed technology may be utilized to enable real-time classification of detected defects in multiple subcomponents of the component in parallel. The component may be, for example, a multi-chip package with silicon interposers, and the subcomponents may include, for example, through-silicon vias and solder joints. Defects in subcomponents of other types of components may be also be classified. One embodiment relates to a method of classifying detected defects in subcomponents of a manufactured component. Another embodiment relates to a product manufactured using a disclosed method of inspecting multiple subcomponents of a component for defects.
1. A method for classifying defects in multiple subcomponents of a component using a high-speed x-ray imaging system, the method comprising, for each subcomponent:
generating a synthetic training set of subcomponent x-ray images using a model of the subcomponent and one or more defect types;
computing feature vectors from the subcomponent x-ray images in the synthetic training set;
transforming the feature vectors to obtain transformed feature vectors;
training a defect classifier using the transformed feature vectors and their corresponding defect types; and
classifying defects in measured high-speed x-ray images using the defect classifier.
2. The method of claim 1 further comprising:
applying the trained defect classifier to defects detected in x-ray images of multiple subcomponents of the component.
3. The method of claim 2 , wherein the defects in the multiple subcomponents are classified in parallel as part of a manufacturing process.
4. The method of claim 2 , wherein the defects in the multiple subcomponents are classified in real-time as part of a manufacturing process.
5. The method of claim 1 , wherein the component comprises a semiconductor package.
6. The method of claim 5 , wherein the subcomponent comprises a through-silicon via or a solder joint.
7. The method of claim 1 , wherein said generating utilizes randomization of defect location and size.
8. The method of claim 7 , wherein said generating further utilizes randomization of defect shape.
9. The method of claim 1 , wherein transforming the feature vectors comprises:
standardizing the feature vectors; and
applying a transformation to transform the feature vectors to a set of transformed feature vectors.
10. The method of claim 1 , wherein the defect classifier comprises a decision-tree based classifier.
11. The method of claim 1 , wherein the defect classifier uses one or more of Hu-moment features, Haar-wavelet features, and Histogram of Oriented Gradients.
12. An apparatus for classifying defects in multiple subcomponents of a component using a high-speed x-ray imaging system, the apparatus comprising:
at least one processor for executing computer-readable code; and
memory for storing and accessing computer-readable code and data;
computer-readable code for generating a synthetic training set of subcomponent x-ray images using a model of the subcomponent and one or more defect types;
computer-readable code for computing feature vectors from the subcomponent x-ray images in the synthetic training set;
computer-readable code for transforming the feature vectors to obtain transformed feature vectors;
computer-readable code for training a defect classifier using the transformed feature vectors and their corresponding defect types; and
computer-readable code for classifying defects in measured high-speed x-ray images using the defect classifier.
13. The apparatus of claim 12 , wherein the method further comprises:
applying the trained defect classifier to defects detected in images of multiple subcomponents of the component.
14. The apparatus of claim 13 , wherein the defects in the multiple subcomponents are classified in parallel as part of a manufacturing process.
15. The apparatus of claim 13 , wherein the method further comprises:
applying the trained defect classifier to defects detected in images of multiple subcomponents of the component.
16. The apparatus of claim 12 , wherein the component comprises a semiconductor package.
17. The apparatus of claim 16 , wherein the subcomponent comprises a through-silicon via or a solder joint.
18. The apparatus of claim 12 , wherein said generating utilizes randomization of defect location and defect size.
19. The apparatus of claim 18 , wherein said generating further utilizes randomization of defect shape.
20. The apparatus of claim 12 , wherein transforming the feature vectors comprises:
standardizing the feature vectors; and
applying a transformation to transform the feature vectors to a set of transformed feature vectors.
21. The apparatus of claim 12 , wherein the defect classifier comprises a decision-tree based classifier.
22. The apparatus of claim 12 , wherein the defect classifier uses one or more of Hu-moment features, Haar-wavelet features, and Histogram of Oriented Gradients.