Seam Pucker Prediction Using Neural Computing

Abstract

Seam pucker can be predicted using thickness, weight, and weft and warp (cantilever) bending stiffness as inputs to a back propagation neural network technique. Correlation coefficients between network approximation and subjective assessment of higher than 0.875 have been reported, which validate the importance of fabric properties used and establish a new prediction technique based on artificial intelligent neural computing. Argues that the integration between the instruments used and the network can provide a new industry tool for combating seam pucker.