Cotton Uniformity Classification By Comprehensive Length Attributes

From the fiber length distribution generated by the dual-beard fibrography, comprehensive length attributes, such as mean length (ML), upper half mean length (UHML), upper quartile length (UQL), 50 percentile span length (50%SpanL), 2.5 percentile span length (2.5%SpanL),  short fiber length (SFC) and uniformity index (UI), can be calculated. Currently, the length uniformity is classified into five categories (very low, intermediate, to very high) by using a single attribute, UI.  But UI does not contain short fiber information, and thus does not realistically reflect the uniformity of cotton. Based on the correlation analysis on these attributes, we found out that ML, SFC and UI are the independent parameters that can be used for cotton length uniformity classification. In this study, we will use the K-means clustering method to determine the natural clusters of the length uniformity based on cotton samples with a wide range of fiber length distributions, and understand the features of the length attributes in each of these clusters.