Study On Accuracy Of Reversing Design Based On ICT

Reverse design is widely used in the design and manufacture of new products.Accuracy research is one of the aspects in reverse design research. Errors in reversedesign are typically comprised of prototype errors, measurement errors, data processingerrors, model errors, manufacturing errors, etc. This paper conducts an accuracyresearch based on the errors in the process of industrial computed tomography (ICT)reverse design.ICT is a technology that measures without damages. The paper draws attention toimage pixel and image scan layer gap, which are factors among the determinants of ICTimage accuracy.Using image enhancement and edge detection during image binarization andalignment in the data processing can improve quality of the image and ensure accuracyon the pixel level. When using marching cubes (MC) algorithm to do meshreconstruction, choosing a suitable threshold plus combining Laplacian smoothing andcurvature smoothing can raise accuracy to an even higher pixel level.With regard to surface modeling, this paper conducted error analyses on a fewcommon NURBS curved surface. The influences of curve direction, starting point, andpoint cloud on the quality of a curved surface are compared through experiment. Amongthese, the accuracy of loft surface is most affected by curve direction. The starting pointcomes second. Block fitting on bounding surface improves the accuracy of curvedsurface. Compared with single surface generation, the quality of curved surface createdthrough a combination of curve, point cloud and mesh fitting is higher.Finally, through appropriate method this paper verified that the accuracy of`ICT-based reverse design is controlled within the layer gap distance, providing someguidance for users with regard to relevant techniques.