Theses and Dissertations


Rui Wang

Issuing Body

Mississippi State University


Thomasson, J. Alex

Committee Member

To, Suminto D.

Committee Member

Jonkman, Jeff

Committee Member

Sui, Ruixiu

Committee Member

Cox, Michael S.

Date of Degree


Document Type

Dissertation - Open Access


Biological Engineering

Degree Name

Doctor of Philosophy


College of Agriculture and Life Sciences


Department of Agricultural and Biological Engineering


Maintaining cotton fiber quality is crucial for the survival of the U.S. cotton industry. Previous studies have indicated that spatial variability of fiber quality parameters exists in cotton fields. If site-specific measurement and prediction of quality is possible, then fiber could be segregated during the harvesting process, thus increasing the overall price a producer would receive for his crop. Because of the importance of fiber micronaire to the textile industry, the fact that micronaire exhibits moderate variation in the field, and the fact that it has been shown to be related to optical properties of cotton fibers, micronaire measurement was considered for quality segregation. Two years? cotton and soil data from two fields in Brooksville, Mississippi were used to investigate how much spatial variation in cotton quality factors could be explained by soil parameters. It was found that spatial variability exists in soil and cotton quality parameters, and micronaire (maturity and fineness) was found to have relatively large variability compared to other quality parameters. About 22 to 35% of the variation in micronaire could be explained by soil parameter variability. Site-specific prediction of micronaire based on only soil seems to be not practical according to the results of this study. Another objective was to develop a methodology for measuring important parameters of cotton crop quality in the field. USDA Micronaire Standard Calibration samples were used in infrared spectral measurements in order to relate their micronaire values to near-infrared and mid-infrared wavelength spectra. Near-infrared reflectance measurements in certain wavelengths ranging from 800 to 2500 nm were found to be closely correlated to micronaire values. Mid-infrared transmittance measurements (ratios) in certain wavelengths ranging from 2.5 to 25 µm were also related to micronaire values. The R2 value of the optimal prediction model was 0.92. This model was validated with HVI measurement of cotton samples from Mississippi and Arizona. Optical sensors based on spectral reflectance and transmittance measurements seem to be a reasonable choice for site-specific harvesting. A practical sensor mounted on cotton picker for measuring cotton micronaire appears to be feasible and a draft design was proposed.



precision agriculture||spatial predication||site-specific harvesting