An intelligent knitted garment defect detection and classification model based on Gabor filter and Modified Elman neural network

In this paper a new knitted garment defect detection and classification model based on 2D Gabor wavelet transform and Elman neural network is introduced. A new modified Elman network is proposed to classify the type of fabric defects which have proportional (P), integral (I), derivative (D) properties. The proposed inspecting model in this study is more feasible and applicable in fabric defect detection and classification. Compared with the traditional back propagation BP network, the successful classification rate obtained by the PID Elman network is higher than the BP neural network with the same number of classification parameter, and training time and classification time used by PID Elman is less than BP neural network.