13.07.2015 Views

1 Using Voxelization and Ray-Tracing to Identify Wall Thinness of ...

1 Using Voxelization and Ray-Tracing to Identify Wall Thinness of ...

1 Using Voxelization and Ray-Tracing to Identify Wall Thinness of ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

32Figure 19: Many voxels are flagged as unsafe when the unsafe cu<strong>to</strong>ff is only one voxelIn many cases, it seemed that highlighting the voxels that were one unit <strong>of</strong> voxel distanceaway from a safe voxel only served <strong>to</strong> emphasize the edges <strong>of</strong> an object. This information isinteresting, but it was turning up <strong>to</strong>o many false positives: voxels that were highlighted as unsafewere, in fact, safe <strong>to</strong> print. Though this did indeed catch overly-narrow areas <strong>of</strong> geometry, weneeded a way <strong>to</strong> reduce the number <strong>of</strong> false positives <strong>to</strong> get the final output more relevant.To try <strong>and</strong> reduce the number <strong>of</strong> false positives, we tried shifting the model slightly <strong>and</strong>running the algorithm again, then comparing the results. Shifting the model in space slightlywhile keeping our voxelization grid centered on the x=0, y=0, z=0 point caused the trianglegeometry <strong>to</strong> fall in<strong>to</strong> different bounding voxels, thus altering the final results. However, overallwe found that this did not produce any significant effect on the number <strong>of</strong> false voxels. Thoughrotating the model or shifting the model would <strong>of</strong>ten eliminate false positives in one area, itwould introduce them in a new area, thereby canceling out any gain. The net effect was roughlythe same, regardless <strong>of</strong> model positioning. Thankfully, from the perspective <strong>of</strong> a 3D printer, wewould rather have false positives than false negatives. If there was a problem area in a model <strong>and</strong>the algorithm did not catch it, that would be greater cause for concern.Another interesting byproduct that occurred when flagging all the distance-1 unsafevoxels was the introduction <strong>of</strong> artifacts. Figure 20 demonstrates this effect.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!