Skip to main content
ARS Home » Research » Publications at this Location » Publication #242007
Title: Applying linear spectral unmixing to airborne hyperspectral imagery for mapping crop yield variability.

Author
item Yang, Chenghai
item Everitt, James
item Bradford, Joe

Submitted to: Proceedings of IEEE International Conference on Image Processing
Publication Type: Proceedings
Publication Acceptance Date: 6/24/2009
Publication Date: 8/26/2009
Citation: Yang, C., Everitt, J.H., Bradford, J.M. 2009. Applying linear spectral unmixing to airborne hyperspectral imagery for mapping crop yield variability. Proceedings of IEEE International Conference on Image Processing. CDROM.

Interpretive Summary: Spectral unmixing techniques can be used to quantify crop canopy cover from remotely sensed imagery and have the potential for mapping the variation in crop yield. This study applied linear spectral unmixing to airborne hyperspectral imagery recorded from a grain sorghum field and a cotton field. Plant cover fractions were positively related to yield and provided better correlations with yield than the majority of the normalized difference vegetation indices. These results indicate that plant cover fraction maps derived from hyperspectral imagery can be used as relative yield maps to characterize crop yield variability.

Technical Abstract: This study evaluated linear spectral unmixing techniques for mapping the variation in crop yield. Both unconstrained and constrained linear spectral unmixing models were applied to airborne hyperspectral imagery recorded from one grain sorghum field and a cotton field. A pair of plant and soil spectra derived from each image was used as endmember spectra to generate unconstrained and constrained plant and soil cover fractions. Yield was positively related to plant fractions and negatively related to soil fractions. For comparison, all 5151 possible narrow-band normalized difference vegetation indices (NDVIs) were calculated from the 102-band images and related to yield. Plant fractions provided better correlations with yield than the majority of the NDVIs. These results indicate that plant cover fraction maps derived from hyperspectral imagery can be used as relative yield maps to characterize crop yield variability.