Skip to main content
ARS Home » Southeast Area » Stoneville, Mississippi » Cotton Ginning Research » Research » Publications at this Location » Publication #249912
Title: Improvement of a Harvester Based, Multispectral, Seed Cotton Fiber Quality Sensor

Author
item SCHIELACK, V. - Texas A&M University
item THOMASSON, J. - Texas A&M University
item Sui, Ruixiu
item MORGAN, C. - Texas A&M University
item HEQUET, E. - Texas Tech University

Submitted to: National Cotton Council Beltwide Cotton Conference
Publication Type: Proceedings
Publication Acceptance Date: 1/15/2010
Publication Date: 4/30/2010
Citation: Schielack, V., Thomasson, J.A., Sui, R., Morgan, C., Hequet, E. 2010. Improvement of a Harvester Based, Multispectral, Seed Cotton Fiber Quality Sensor. National Cotton Council Beltwide Cotton Conference. CD ROM p. 596-604.

Interpretive Summary: The profitability of cotton is dependent upon two main factors, yield and quality. Cotton yield monitor can be used to provide a yield map of cotton field. This yield map can be used to adjust field management practices to help increase total yield throughout the cotton field, but it contains no information about the quality of cotton that has been picked. If a quality map of the field were also available, adjustments in field management could help each field produce not only as much cotton as possible, but also the best quality cotton possible. A multispectral sensor for in-situ seed cotton fiber quality measurement was developed and tested at Texas A&M University. Results of initial testing of the sensor using machine harvested seed cotton have shown promise. Improvements have been made to the sensor and the measurement method to meet the objectives of improving the accuracy and repeatability of cotton micronaire measurement. The sensor will take five images of a sample of seed cotton - two images in wavebands sensitive to trash content in cotton and three images in wavebands sensitive to micronaire of the cotton. Using the two images sensitive to trash, pixels that represent trash in the image will be identified and excluded from further processing. Using the remaining pixels as a map of cotton lint, the images sensitive to micronaire will be processed to determine micronaire value of the cotton. The improved sensor for seed cotton fiber quality measurement is under lab test.

Technical Abstract: A multispectral sensor for in-situ seed cotton fiber quality measurement was developed and tested at Texas A&M University. Results of initial testing of the sensor using machine harvested seed cotton have shown promise. Improvements have been made to the system and the measurement method to meet the objectives of improving the accuracy and repeatability of micronaire values from the sensor when compared to the quality measured by standard AFIS and HVI machines. The sensor will take five images of a sample of seed cotton - two images in wavebands sensitive to harvester trash and three images in wavebands sensitive to micronaire. Using the two images sensitive to trash, pixels that represent trash in the image will be identified and excluded from further processing. Using the remaining pixels as a map of cotton lint, the images sensitive to micronaire will be processed and the micronaire value compared to standard measurements. As of the writing of this paper, improvements have been made to the system, but testing is not yet completed.