Measurement of short fiber contents in raw cotton using dual-beard images

Abstract

The short fiber content (SFC) in raw cotton is an extremely important attribute in the modern cotton classification system, because of its impact on the quality of yarn manufacturing. However, the current SFC measuring methods are costly, time-consuming and tedious for accurate and quick testing. This study aimed to develop a new way to calculate a weight-based short fiber content ( SFCw) from a dual-beard specimen via image-processing and bell-shaped fibrogram modeling techniques. A sample from a cotton sliver was clamped and combed in the opposite directions to form two tapered fringes, and then scanned on a desktop scanner to generate a dual-beard image (DBI). The grayscale values in each column of the DBI were accumulated to derive a bell-shaped fibrogram representing the fiber quantity distributions over the distance from the fibrogram center. Based on the fibrogram, a special hierarchical model was proposed to calculate SFCw in two different length limits—12.7 and 16 mm—defined respectively in the USA and China’s specifications. Fifteen samples were tested to compare the SFCw results from the DBIs to those from the Advanced Fiber Information System (AFIS PRO2). Significant correlations were found between the SFCw values from DBIs and AFIS. The regressive equations for SFCw prediction were validated with six additional samples. As shown in the Bland–Altman plots, a good agreement existed between the two sets of SFCw values, demonstrating the DBI’s potential to be an accurate, rapid and portable approach for measuring SFCw, and to transform the SFCw testing from a dedicated machine to an off-the-shelf scanner.