Classifying cotton bark and grass extraneous matter using image analysis

Authors: Whitelock, Derek; Hughs, Sidney; Armijo, Carlos

Submitted to: Textile Research Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/29/2016
Publication Date: 5/1/2017
Citation: Whitelock, D.P., Hughs, S.E., Armijo, C.B. 2017. Classifying cotton bark and grass extraneous matter using image analysis. Textile Research Journal. 87(8):891-901.
DOI: https://doi.org/10.1177/0040517516641360

Interpretive Summary: Most the cotton grown in the U.S. is classed by the USDA Agricultural Marketing Service and the price of U.S. cotton is based on that classification. This classification is determined using high volume instrument machine measurements for all attributes with the exception of extraneous matter (EM - any substance in cotton other than fiber or leaf) and some special conditions, which are determined manually by a human classer. In order to develop a machine EM classing system, a better understanding of what triggers a human classer EM call is needed. The goal of this analysis was to develop criteria for identifying bark/grass EM objects in cotton samples using machine measurements. In this analysis, human classers identified bark/grass EM objects in samples of cotton. Images of those samples were then analyzed to make shape and color measurements of the bark/grass objects and other leaf particles. Results showed that there were very significant differences in shape and color between bark/grass objects and leaf particles. To classify items in a sample, a mathematical model with shape, color, and specially transformed shape variables worked best; correctly classifying about 99% of items. These results will aid in developing new machine classing systems for EM that accurately represent the “gold standard” human classer.

Technical Abstract: Cotton extraneous matter (EM) and special conditions are the only cotton quality attributes still determined manually by USDA-AMS classers. To develop a machine EM classing system, a better understanding of what triggers a classer EM call is needed. The goal of this work was to develop new information about cotton EM, such as bark and grass, and leaf particles, using machine measurements, to aid in the development of instrumentation for cotton quality measurements. AMS classers were tasked in identifying and denoting bark/grass in large-area, color images of cotton samples. Image segmentation analysis was applied to detect non-cotton items, such as leaf particles, and the classer denoted bark/grass objects were segmented manually. Further image analysis was used to measure shape and color parameters of these bark/grass objects and leaf particles in the sample images. These measurements of the bark/grass objects and leaf particles were compared and logistical regression analyses conducted to evaluate classification. For every shape and color parameter, there were significant differences between the bark/grass objects and the detected leaf particles in the images. The differences were greater for the shape parameters than for the color parameters. A classification model with shape, color, and log-transformed shape parameters consistently classified the bark/grass objects and leaf particles most accurately with 99.5 and 97.6% correct classification rate, respectively. However, classification models that were 99% correct classifying manually segmented bark/grass were only about 77% correct when applied to the machine detected bark/grass particles.