Identification of cashmere and wool based on LBP and GLCM texture feature selection

Abstract

There are invalid and redundant features in the texture feature extraction method of cashmere and wool fibers, which leads to the low recognition accuracy. In this paper, a novel texture feature selection method based on local binary pattern, the gray level co-occurrence matrix algorithm and chi-square test was proposed to sufficiently extract the effective features of these two fibers. Firstly, the collected images of cashmere and wool fibers are processed to obtain the clear texture images with background removed by pre-processing algorithm and local binary pattern. Then, the texture features are calculated by the gray level co-occurrence matrix, and the optimal 5-dimensional features are extracted by chi-square test to represent the texture information of cashmere and wool. Finally, the two fibers are automatically classified and recognized based on the support vector machine. The experimental results show that the proposed method obtained a high recognition accuracy with the percent of 94.39. It verifies that the method based on texture feature selection is effective to identify cashmere and wool fibers.

A weakly supervised learning pipeline for profiled fibre inspection

“…The Otsu algorithm [44] is used to estimate the optimal threshold for binary fibre segmentation. To remove noise and smooth contour of the segmented fibre, the morphological algorithm [8], opening operation is employed. The cross session of the fibre instance in a patch is segmented as the connected region, based on which the contour is extracted through the Candy edge algorithm [45].…”

“…There are two main paradigms for fibre detection: traditional image processing-based methods and deep learning-based methods. Traditional image processing-based methods [17] segmented fibres by its spatial statistical features and could be classified into four types, threshold algorithms [4,18], morphological algorithms [8], region-based algorithms [19,20], and edge detection-based algorithms [21,22]. Most of these methods are proposed to deal with fibre overlapping, adhesion and breakage which are the main challenges for fibre detection.…”

“…Since spectral characteristics, e.g., light absorptions in variance with wavelengths can be reflected by the spectral signatures sampled by a near-infrared spectroscopy, they have been exploited in several works [24,25]. Spatial features such as geometric features [26,27] including fibre diameter and scale thickness, texture features [8,28,29] including LBP texture and wavelet texture and composite features which combine several features, are also commonly used in the fibre feature extraction. As for the classification stage, classic pattern recognition methods are mostly used, e.g., decision tree, Bayesian model and K-Means clustering.…”

“…With the advent of computer vision technology, a variety of methods have been proposed to automate profiled fibre detection [4, 5], classification [6–8] and shape factor estimation [9, 10]. Recently, many works focus on the recognition task which integrates detection and classification in a network.…”