Animal fiber recognition based on feature fusion of the maximum inter-class variance

Abstract Cashmere and wool are common raw materials in the textile industry. The clothes made of cashmere popular because excellent comfort. A system that can quickly automatically classify two will improve efficiency fiber recognition We propose a classification method fibers based on feature fusion using maximum inter-class variance. First, target area is obtained by preprocessing algorithm. Second, features sub-images extracted through algorithm Discrete Wavelet Transform. It linearly fused introducing weight approximate detailed features. separability data be achieved Finally, different classifiers used to evaluate performance proposed method. support vector machine classifier achieve highest rate, with an accuracy 95.20%. experimental results show rate vectors improved 6.73% compared original describing image. verifies provides effective solution for automatic wool.