Submitted to: International Journal of Agricultural and Biological Engineering
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: August 28, 2009
Publication Date: September 30, 2009
Citation: Zhang, H., Lan, Y., Lacey, R., Hoffmann, W.C., Huang, Y. 2009. Analysis of vegetation indices derived from aerial multispectral and ground hyperspectral data. International Journal of Agricultural and Biological Engineering. 2:1-8. Interpretive Summary: New methods are needed to incorporate remote sensing along with Global Positioning Systems, Geographic Information Systems, and variable-rate spray technology to help farmers maximize the economic and environmental benefits of crop management through precision agriculture. An automated four-channel multispectral camera was coupled with a specifically designed camera control system in an aircraft to consistently produce precise aerial images for the spatial analysis of crop conditions; a ground-based 128-channel hyperspectral instrument simultaneously monitored the crop conditions. Vegetation indices based on aerial multispectral and ground-based hyperspectral measurements yielded significant differences in monitoring the general growth status of cotton and soybean. The systems and techniques developed here will enhance the use of aerially-acquired images and ground truth measurements for precision crop management.
Technical Abstract: Aerial multispectral images are a good source of crop, soil, and ground coverage information. Spectral reflectance indices provide a useful tool for monitoring crop growing status. A series of aerial images were acquired by an airborne MS4100 multispectral imaging system on the cotton and soybean field. Ground truth hyperspectral data were acquired with a Fieldspec spectroradiometer at the same periods. The Normalized Difference Vegetative Index (NDVI), Simple Ratio (SR), and Soil Adjusted Vegetation Index (SAVI) were calculated from both systems and compared. The information derived from aerial multispectral images has shown the potential to monitor the general growth status of crop field. The vegetation indices acquired using both systems were sensitive to changes in the crop field. The mean SAVI value was significantly different from that of NDVI from aerial images for both fields (p-value is 0.0334 for soybean and 0.0537 for cotton field at alpha=0.05 and 0.1 levels, respectively). The correlation coefficient of the NDVI values from both systems was 0.439 for soybean field and higher than the correlation coefficients of the SR and SAVI values.