Multispectral Sensor for In-Situ Cotton Fiber Quality Measurement

Abstract

Reflectance spectra of cotton fiber samples having different fiber quality levels were measured with a high-resolution spectrophotometer. Reflectance spectra of the cotton samples were processed with waveband averaging and wavelet analysis, and then related to micronaire by using multiple linear regression. Regression models indicated that the micronaire had a strong correlation (r2 = 0.89) with the reflectance values at seven 100-nm wavebands (1120, 1296, 1550, 1664, 1852, 2020, and 2340 nm). In wavelet analysis, six wavelet regressors were identified and entered into the regression model. These models also indicated a very strong correlation between micronaire and reflectance spectra in the wavelet domain (r2 = 0.97). A prototype of cotton fiber quality sensor was developed based on the characteristics of the cotton fiber reflectance spectrum. The sensor consists of a VisGaAs camera, optical bandpass filters, halogen light source, and an image collection and process system. The sensor was tested in the laboratory conditions. Images of lint samples at three near infrared (NIR) wavebands (1450, 1550, and 1600 nm) were acquired and analyzed to determine the relationship between the image pixel value and cotton fiber micronaire. Results showed that the sensor was capable of accurately assessing the fiber micronaire (r2 = 0.99). This sensor could be used for measuring cotton fiber quality along with their corresponding spatial data from GPS as cotton is harvested in fields, which makes it possible to generate cotton fiber quality maps. It also has the potential being used for segregating cotton based on fiber quality during harvesting.