Inspection system for three-dimensional printer and methods thereof

1. A method of inspecting drop ejection in a printing system, comprising:

capturing an image of each of a plurality of drops of a print material after ejection from an ejector of a printing system;

creating a temporally averaged image from each image of the plurality of drops of print material; and

classifying one of the plurality of drops of print material based on the temporally averaged image that was created.

2. The method of claim 1 , further comprising using a pretrained convolutional neural network for classifying one of the plurality of drops.

3. The method of claim 1 , further comprising comparing the temporally averaged image to another temporally averaged image to classify one of the plurality of drops.

4. The method of claim 1 , further comprising extracting a region of interest (ROI) from the image of each of the plurality of drops of print material.

5. The method of claim 4 , wherein extraneous image data is reduced from one or more areas of the image surrounding each of the plurality of drops of print material.

6. The method of claim 1 , further comprising extracting a region of interest (ROI) from the temporally averaged image from each image of the plurality of drops of print material.

7. The method of claim 6 , wherein extraneous image data is reduced from one or more areas of the temporally averaged image surrounding each image of the plurality of drops of print material.

8. The method of claim 1 , further comprising capturing of the image of each of a plurality of drops of print material after ejection by from about 1 ms to about 200 ms.

9. The method of claim 1 , further comprising illuminating a space between the ejector and a substrate where a three-dimensional part is being built.

10. The method of claim 9 , wherein the illuminating is done continuously.

11. The method of claim 1 , further comprising:

indicating an unstable jetting result;

pausing an operation of the printing system; and

performing an intervention on the printing system.

12. The method of claim 1 , wherein the print material comprises a metal, a metallic alloy, or a combination thereof.

13. The method of claim 1 , wherein the print material comprises aluminum.

14. A method of inspecting drop ejection in a printing system, comprising:

capturing an image of each of a plurality of drops of print material after ejection from an ejector of a printing system;

creating a temporally averaged image from each image of the plurality of drops of print material;

extracting a region of interest (ROI) from each temporally averaged image from each image of each of the plurality of drops of print material;

comparing the region of interest (ROI) from a first temporally averaged image to a region of interest (ROI) from a second temporally averaged image to generate a jetting result; and

classifying the jetting result of one of the plurality of drops of print material.

15. The method of claim 14 , further comprising:

indicating an unstable jetting result;

pausing an operation of the printing system; and

performing an intervention on the printing system.

16. The method of claim 14 , further comprising using a pretrained convolutional neural network for classifying the jetting result of one of the plurality of drops of print material.

17. The method of claim 14 , further comprising capturing of the image of each of a plurality of drops of print material after ejection by from about 1 ms to about 200 ms.

18. The method of claim 14 , further comprising continuously illuminating a space between the ejector and a substrate where a three-dimensional part is being built.

19. The method of claim 14 , wherein extraneous image data is reduced from one or more areas of each image of the plurality of drops of print material.

20. A printing system, comprising:

an ejector for jetting a print material, comprising:

a structure defining an inner cavity; and

a nozzle orifice in connection with the inner cavity and configured to pulse to eject one or more drops of liquid print material;

a camera comprising a high-speed shutter wherein the shutter is synchronized to an ejector pulse; and

a video analytic framework coupled to the ejector and the camera configured to generate a jetting result for each of the one or more drops of liquid print material.

21. The printing system of claim 20 , further comprising a substrate for receiving one or more drops of liquid print material.

22. The printing system of claim 21 , wherein the camera is located in a plane between the ejector and the substrate.

23. The printing system of claim 20 , further comprising an electronic circuit connecting the camera to the ejector and configured to trigger the camera after a pulse to eject one or more drops of liquid print material.

24. The printing system of claim 20 , wherein the camera is triggered 1 ms to about 200 ms after the pulse to eject one or more drops of liquid print material.

25. The printing system of claim 20 , wherein the camera further comprises a long-length focal lens.

26. The printing system of claim 21 , further comprising a continuous illumination source configured to illuminate a space between the ejector and the substrate.

27. The printing system of claim 20 , wherein the video analytic framework produces an image extraction to generate a jetting result for each of the one or more drops of liquid print material.

28. The printing system of claim 27 , wherein the image extraction is a temporal average of more than one image.

29. The printing system of claim 20 , wherein the print material comprises a metal, a metallic alloy, or a combination thereof.