Object manufacturing visualization

1. A method for object manufacturing visualization by an electronic device, comprising:

determining, using a machine learning model, a predicted thermal image of a current layer in additive manufacturing, wherein the predicted thermal image is determined by the machine learning model based on at least using a contone map that corresponds to the current layer and a captured thermal image of a previous layer;

obtaining a captured thermal image of the current layer; and

presenting, via graphical user interface, a graphical overlay of the predicted thermal image of the current layer superimposed the captured thermal image of the current layer.

2. The method of claim 1 , further comprising stacking a plurality of predicted thermal images, wherein the graphical overlay is presented with the plurality of predicted thermal images.

3. The method of claim 1 , further comprising masking the predicted thermal image and the captured thermal image using a contone map to produce part maps.

4. The method of claim 1 , further comprising:

calculating a difference between the predicted thermal image and the captured thermal image;

calculating an average of the difference; and

calculating an anomaly score based on the average and based on statistical values of the machine learning model, wherein the anomaly score is calculated by subtracting a mean of the machine learning model from the difference calculated, and further dividing by a standard deviation of the machine learning model.

5. The method of claim 1 , further comprising:

calculating a set of difference values between the predicted thermal image and the captured thermal image; and

calculating a set of anomaly scores based on the set of difference values and based on statistical values of the machine learning model.

6. The method of claim 5 , further comprising comparing the set of anomaly scores to a set of thresholds to produce a set of grades over the captured thermal image.

7. The method of claim 6 , further comprising color coding the graphical overlay based on the set of grades.

8. The method of claim 6 , wherein the set of thresholds comprises multiples of a standard deviation of the machine learning model.

9. The method of claim 8 , wherein an anomaly score that is more than two standard deviations below a mean of the machine learning model indicates an under-melted portion of an object.

10. The method of claim 8 , wherein an anomaly score that is more than two standard deviations above a mean of the machine learning model indicates an over-melted portion of an object.

11. A three-dimensional (3D) printing device, comprising:

a print head to print an agent based on a contone map;

a thermal projector;

a thermal sensor; and

a controller, wherein the controller is to:

receive a captured thermal image from the thermal sensor for a current layer;

calculate a predicted thermal image for a current layer using a neural network based on a plurality of contone maps for a current layer and a captured thermal image of a previous layer; and

present a graphical overlay that superimposes the predicted thermal image and the captured thermal image of the current layer.

12. The 3D printing device of claim 11 , wherein the plurality of contone maps are fusing contone maps.

13. The 3D printing device of claim 12 , wherein the plurality of contone maps are detailing contone map.

14. A non-transitory tangible computer-readable medium storing executable code, comprising:

code to cause a processor to obtain fusing contone maps of layers and detailing contone maps of layers;

code to cause the processor to obtain captured thermal images of layers;

code to cause the processor to calculate predicted thermal images for the layers using a neural network based on the fusing contone maps, the detailing contone maps, and the captured thermal images for the corresponding layers; and

code to cause the processor to present a first viewport of stacked fusing contone maps, a second viewport of stacked detailing contone maps, a third viewport of stacked predicted thermal images, and a fourth viewport of stacked captured thermal images.

15. The computer-readable medium of claim 14 , wherein the stacked fusing contone maps, the stacked detailing contone maps, the stacked predicted thermal images, and the stacked captured thermal images exclude a non-object portion.