Method and apparatus for rapid inspection of subcomponents of manufactured component
The presently-disclosed technology enables real-time inspection of a multitude of subcomponents of a component in parallel. For example, the component may be a semiconductor package, and the subcomponents may include through-silicon vias. One embodiment relates to a method for inspecting multiple subcomponents of a component for defects, the method comprising, for each subcomponent undergoing defect detection: extracting a subcomponent image from image data of the component; computing a transformed feature vector from the subcomponent image; computing pairwise distances from the transformed feature vector to each transformed feature vector in a training set; determining a proximity metric using said pairwise distances; and comparing the proximity metric against a proximity threshold to detect a defect in the subcomponent. Another embodiment relates to a product manufactured using a disclosed method of inspecting multiple subcomponents of a component for defects. Other embodiments, aspects and features are also disclosed.
1. A method for inspecting interior features corresponding to multiple subcomponents of a component for defects, the method comprising, for each subcomponent undergoing defect detection:
inspecting an interior feature corresponding to a subcomponent of a component with an x-ray imaging system during x-ray imaging system operation;
creating x-ray image data of the component during x-ray imaging system operation;
extracting a subcomponent image corresponding to the interior feature from the x-ray image data of the component;
computing a transformed feature vector from the subcomponent image;
computing pairwise distances from the transformed feature vector to each transformed feature vector in a training set;
determining a proximity metric using said pairwise distances;
comparing the proximity metric against a proximity threshold to detect a defect in the subcomponent;
the transformed feature vector corresponds to a characteristic feature of submicron size; and
the method simultaneously evaluates the multiple subcomponents during parallel procedural steps.
2. The method of claim 1 , wherein the component comprises a semiconductor package.
3. The method of claim 2 , wherein the subcomponent comprises a through-silicon via.
4. The method of claim 1 , wherein computing the transformed feature vector comprises:
computing a feature vector; and
applying a linear transformation to transform the feature vector to a transformed feature vector.
5. The method of claim 1 , wherein computing the pairwise distances comprises computing dot products between pairs of the transformed feature vectors.
6. The method of claim 1 , wherein determining the proximity metric comprises determining a pairwise distance at a percentile level in a distribution of the pairwise distances.
7. The method of claim 1 , wherein the proximity threshold is determined from a distribution of proximity metrics, and wherein each proximity metric in said distribution corresponds to a transformed feature vector in the training set.
8. The method of claim 1 , wherein the defects are unclassified upon detection.
9. The method of claim 1 , wherein the multiple subcomponents are inspected in real-time as part of a manufacturing process.
10. A product manufactured using a process that involves a method for simultaneous inspection of multiple subcomponents of a component for defects, the method comprising, inspecting the multiple subcomponents with an x-ray inspection device, with the inspecting the multiple subcomponents performed in parallel in real-time for each subcomponent undergoing defect detection:
extracting a subcomponent image from image data of the component;
computing a transformed feature vector from the subcomponent image;
computing pairwise distances from the transformed feature vector to each transformed feature vector in a training set;
determining a proximity metric using said pairwise distances;
comparing the proximity metric against a proximity threshold to detect a defect in the subcomponent;
the transformed feature vector corresponds to a characteristic feature of submicron size; and
the process simultaneously evaluates the multiple subcomponents during parallel procedural steps.
11. The product of claim 10 , wherein the component comprises a semiconductor package.
12. The product of claim 11 , wherein the subcomponent comprises a through-silicon via.
13. The product of claim 10 , wherein computing the transformed feature vector comprises:
computing a feature vector; and
applying a linear transformation to transform the feature vector to a transformed feature vector.
14. The product of claim 10 , wherein computing the pairwise distances comprises computing dot products between pairs of the transformed feature vectors.
15. The product of claim 10 , wherein determining the proximity metric comprises determining a pairwise distance at a percentile level in a distribution of the pairwise distances.
16. The product of claim 10 , wherein the proximity threshold is determined from a distribution of proximity metrics, and wherein each proximity metric in said distribution corresponds to a transformed feature vector in the training set.
17. The product of claim 10 , wherein the defects are unclassified upon detection.
18. The product of claim 10 , wherein the multiple subcomponents are inspected in real-time as part of a manufacturing process.
19. A system for inspecting a subcomponent of a semiconductor component for defects, the system comprising:
an x-ray imaging apparatus for obtaining image data of the subcomponent of the semiconductor component;
at least one processor for executing computer-readable code; and
memory for storing and accessing computer-readable code and data;
computer-readable code for extracting a subcomponent image from the image data of the component;
computer-readable code for computing a transformed feature vector from the subcomponent image;
computer-readable code for computing pairwise distances from the transformed feature vector to each transformed feature vector in a training set;
computer-readable code for determining a proximity metric using said pairwise distances;
computer-readable code for comparing the proximity metric against a proximity threshold to detect a defect in the subcomponent;
the transformed feature vector corresponds to a characteristic feature of submicron size;
the subcomponent is one of multiple subcomponents; and
the system simultaneously evaluates the multiple subcomponents during parallel procedural steps.