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.

Preliminary Study on Nitrogen Monitoring of Potato Plants Using Multispectral Image

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan

Citation:  Paper number  131620028,  2013 Kansas City, Missouri, July 21 - July 24, 2013. (doi: @2013
Authors:   Hong Sun, Hong Sun, Minzan Li, Qin Zhang, K. Alva Ashok, Zhiyan Zhou
Keywords:   Multi-spectral image vegetation index nitrogen content potato monitoring

Abstract. In order to explored a feasible method to monitor the nitrogen content of potato in the field. The spectrum and multi-spectral imagery techniques were applied. A portable spectrometer and a MS4100 multispectral camera were used to collect spectral reflectance and multi-spectral image in the field. The experiments were conducted under three different fertilizer treatments with Low, High and Meddle Nitrogen (N) availability. The leave of potato plants were sampled for N analyses. The image of potato canopy was processed following the image calibration, image segmentation, and image parameter calculation. Each channel of multi-spectral image was used to calculate the detecting parameters including blue (BIA), green (GIA), red (RIA) and near-infrared (NIRIA). Meanwhile, the vegetation indices (DVI, RVI, NDVI, SAVI, et al.) widely used in remote sensing were selected as the parameters for nitrogen monitoring of potato in the field. Then, the correlation between image parameters and nitrogen content of potato plant was analyzed under different fertilizer treatments and different growth stages. It was showed that the middle treatment was best level to evaluate the nitrogen content using multispectral image, and it was sensitive to evaluate the nitrogen content at maturity stage. The PLS models were built to estimate the nitrogen content of potato. The results indicated that when NIRIA, RIA and RVI were applied, the predicting r2 was 0.30; When NIRIA, RIA, RVI, NDVI and SAVI were used, the predicting r2 was 0.35.

(Download PDF)    (Export to EndNotes)