Snippet Counting for Cotton Length Distribution Measurement Using Image Analysis

Abstract

This paper presents a new method of measuring the number-length distribution of cotton fibers using a snippet-counting method and image analysis techniques. The test samples were aligned at the baseline and combed to form a tuft web, or beard, which was cut into a number of segments of a specific length by a fiber cutter. The snippets in different cut segments were collected and distributed on separate black glossy papers, and then they were digitized and scanned in order into a computer for image analysis. The scanning order indicated the distance of each segment from the cutting baseline, and the number of the snippets on a glossy paper was the frequency of fibers present in the current segment. A specially designed image analysis program was used to trace all snippets in a scanned image, so that the number of snippets could be calculated. After all of the images were analyzed, the fiber numbers in different segments could be summarized, and from these number-length data the distribution, maximum length, mean length, and other fiber statistics could be derived.