Objective evaluation of seam pucker using artificial intelligence. Part II: Method of evaluating seam pucker

Abstract

A new quantitative method to evaluate seam pucker with five shape parameters is proposed using three-dimensional image analysis and artificial intelligence. The shape parameters include the number of wave generating points, the wave amplitudes, and the wavelengths on the line next to the seam and on the edge line. Measured shapes of puckered fabrics are converted into numerical data in the three-dimensional coordinate system, and the data are transformed into power spectra using fast Fourier transformation. Also, artificial intelligence techniques, including neural networks and fuzzy logic (or neurofuzzy algorithm), are used to recognize the characteristics of seam pucker. To obtain better quantitative evaluations of seam pucker, 300 neurofuzzy engines are constructed and trained using reference puckered shapes produced by the simulator described in Part I. Power spectra of the measured data are transformed into specified fuzzy patterns through the fuzzification process to train the neurofuzzy engines. The five shape parameters of puckered shapes are obtained from pattern recognition by the trained neurofuzzy engines through the defuzzification process.