Method of correcting a three-dimensional image of an electronic circuit

1. In an optical inspection system comprising an image processing computer system connected to an image sensor, a method of determining condition of a deformed object comprising:

providing a reference model of the deformed object comprising first areas;

determining in the initial three-dimensional image second areas corresponding to the first areas;

determining a first geometric transformation which matches the second areas with the first areas;

determining a corrected three-dimensional image based on the first geometric transformation and on the initial three-dimensional image;

providing the reference model of the deformed object comprising third areas;

acquiring at least one two-dimensional image of the object with the image sensor;

searching in the two-dimensional image for at least fourth areas corresponding to the third areas;

determining a second geometric transformation which matches the fourth areas with the third areas;

applying the second transformation to the first areas to obtain fifth areas the two-dimensional image;

searching in the initial three-dimensional image for the second areas corresponding to the fifth areas;

determining a third geometric transformation which matches the second areas with the fifth areas;

determining the corrected three-dimensional image based on the second and third geometric transformations and on the initial three-dimensional image; and

determining condition of the deformed object based on the corrected three-dimensional image.

2. The method of claim 1 , wherein the object comprises a printed circuit comprising at least a support, conductive tracks on the support, and a resin layer at least partly covering the conductive tracks.

3. The method of claim 2 , wherein at least one of the first areas comprises a planar portion of the resin layer.

4. The method of claim 2 , wherein at least one of the first areas comprises the edge of a conductive track.

5. The method of claim 2 , wherein at least one of the third area comprises a pattern of an opening crossing the printed circuit.

6. The method of claim 2 , wherein at least one of the third areas comprises a pattern of a conductive track changing direction.

7. The method of claim 1 , comprising a step of providing an initial model which does not comprise the first and third areas and a step of modifying the initial model to add thereto a description of the first and third areas.

8. The method of claim 7 , comprising the steps of, in a training phase:

acquiring a three-dimensional image of a training object;

correcting the deformations of the three-dimensional image of the training object;

determining first and third areas on the corrected three-dimensional image of the training object; and

modifying the initial model to add thereto a description of the determined first and third areas.

9. The method of claim 8 , wherein the reference model is representative of a two-dimensional image.

10. The method of claim 8 , further comprising determining a fourth geometric transformation which matches the points of the three-dimensional image of the training object with the pixels of the initial model.

11. The method of claim 1 , comprising the steps of:

determining a first three-dimensional image of an empty printed circuit;

determining the first geometric transformation associated with the first three-dimensional image;

correcting the first three-dimensional image from the first geometric transformation associated with the first three-dimensional image;

determining a second three-dimensional image of a printed circuit covered with welding paste blocks;

determining the first geometric transformation associated with the second three-dimensional image;

correcting the second three-dimensional image based on the first geometric transformation associated with the second three-dimensional image; and

comparing the first and second corrected three-dimensional images.

12. The method of claim 11 , comprising a step of selecting first areas from among said first areas for the determination of the first geometric transformation associated with the second three-dimensional image.

13. The method of claim 1 , comprising a step of selecting third areas from among said third areas for the determination of the first geometric transformation associated with the second three-dimensional image.