Submitted to: Geombinatorics Journal
Publication Type: Peer reviewed journal
Publication Acceptance Date: 9/10/2000
Publication Date: 10/15/2000
Citation: Interpretive Summary: The economic impact of the research reported here is a direct result of ginners', merchants', and finally textile mills' ability to now identify the amount of different types of trash in their lay down. For instance, a textile mill can steer clear of cotton bales that could hurt production of a given fabric, improving throughput and lessening rejected product. This paper discusses using geometric methods to identify trash, (non-lint, non-fiber material) in ginned cotton. This paper has shown the author's ability to identify four types of trash: bark1 (stringy), bark 2 (stick-like), leaf, and pepper.
Technical Abstract: This paper discusses the use of a geometric approach to classify different types of trash (non-lint, non-fiber material) in ginned cotton. Pieces of trash can have complicated shapes, and a good approximating family of sets is required. Which approximation family is the best? The corresonding optimization problem is reduced to a geometric one: namely, under reasonable conditions, an optimal family must be shift-, rotation-, and scale-invariant. This geometric reduction indicates that the best approximating low-dimensional families consist of sets with the existing empirical classification of trash into bark1, bark2, leaf, and papper trash.