Data structures and algorithms for precise defect location by analyzing artifacts
Described are methods and systems for providing improved defect detection and analysis using infrared thermography. Test vectors heat features of a device under test to produce thermal characteristics useful in identifying defects. The test vectors are timed to enhance the thermal contrast between defects and the surrounding features, enabling IR imaging equipment to acquire improved thermographic images. In some embodiments, a combination of AC and DC test vectors maximize power transfer to expedite heating, and therefore testing. Mathematical transformations applied to the improved images further enhance defect detection and analysis. Some defects produce image artifacts, or “defect artifacts,” that obscure the defects, rendering difficult the task of defect location. Some embodiments employ defect-location algorithms that analyze defect artifacts to precisely locate corresponding defects.
1. An imaging system for localizing a defect on an object under test, the imaging system comprising:
an excitation source for applying a test vector to the defect, wherein the defect exhibits a response to the applied test vector;
an image detector positioned to receive the response and produce an image therefrom, the image including the defect and a defect-artifact encompassing a location of the defect; and
means for classifying the defect-artifact as a specific defect-artifact type, defect-artifact type classifications including a point-type, a line-type, and a corner-type;
means for applying an algorithm to the image to locate the defect within the defect-artifact, the algorithm being adapted for the specific defect-artifact type;
means for iteratively applying a first morphological filtering algorithm to the image for each of i threshold levels, where i is an integer greater than 1, thereby producing a set of i thresholded images wherein the set of i thresholded images includes the defect-artifact, and
means for generating a data structure associating the type-specific defect-artifacts from different type-specific images, the data structure linking a tree of thresholded defect-artifact images using successively greater threshold levels for gradient peak profiling.
2. The imaging system of claim 1 , wherein the response is a thermal response.
3. The imaging system of claim 1 , further comprising a means for applying a second morphological filtering algorithm to each of the i thresholded images, the second morphological filtering algorithm removing all types of defect-artifacts except for the specific defect-artifact type.
4. The system of claim 1 wherein the gradient peak profiling includes a means for determining a maxima of a gradient in response to the threshold levels.
5. An inspection system comprising:
an excitation source that excites an object under test object in a way that produces an image response in an area of the object containing a defect;
an infrared imaging device to capture the image response, the image response including a defect response representative of the defect and a defect-artifact;
an image-processing means for classifying the defect-artifact as a specific defect-artifact type, defect-artifact type classifications including a point-type, a line-type, and a corner-type, and for processing the image response in accordance with an algorithm to locate the defect, the algorithm being adapted for the specific defect-artifact type;
an image-processing means for applying a plurality of threshold values to the image response to produce a plurality of thresholded images, wherein a set of the thresholded images includes the defect-artifact, and for filtering each thresholded image to remove all of the defect-artifacts except the specific defect-artifact type, and
an image-processing means for calculating a gradient between a pair of threshold values of the defect-artifact of the specific defect-artifact data type, and for using the gradient to locate the defect in the area of the object.
6. The system of claim 5 further comprising a means for determining a maxima of the gradient in response to the threshold values.