Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

The paper addresses a practical problem associated with using hyperspectral imaging technique for whole wheat assessment. The curved surface of wheat kernels introduces spectral variability among spectral data from different parts of the kernels. Two functions viz. simulated ellipsoidal surface fitting and morphological shrinking algorithm, were used to compensate for reflectance variation caused by the curved surface of wheat kernels. Kernels of four different western Canadian wheat classes were tested and results showed that morphological shrinking method was effective in reducing the overall spectral variability. It was also found that ellipsoidal surface fitting was as effective as morphological shrinking method along the minor axis direction of wheat kernel surface but failed to reduce variation along the major axis. The coefficients of variation (CV) values were reduced to approximately half of the original values along the minor axis direction within the 1100 nm - 1350 nm spectral region after applying both methods.