Multispectral sensor for detecting nitrogen in corn plants.

Abstract

Current heavy reliance on agricultural chemicals in US agriculture has raised many environmental and economic concerns. Farmers need to find a way of optimal application of agricultural chemicals. Among agricultural chemicals, nitrogen (N) is the most important and essential for growing crops. In this research, a nitrogen sensor was developed to assess nitrogen status of corn plants and to apply fertilizers precisely. An in-field hyperspectral sensor system was successfully constructed for real-time data acquisition and was tested in a commercial corn field, however it was found that the wide flood lighting system had greater impact on the sensor measurement than the variation of nitrogen itself. This research showed that the 1 derivative analysis algorithm for 1999 data set measured by a spectrophotometer produced good predictability to N (%) content. INTRODUCTION Current heavy reliance on agricultural chemicals in US agriculture has raised many environmental and economic concerns. For an effective use of agricultural chemicals to get maximum yield without raising any environmental concern, farmers need to find a way of applying optimum amount of agricultural chemicals. Among agricultural chemicals, nitrogen (N) is the most important and essential for growing crops and is also most concerning nutrient element for maintaining healthy environment. Its accurate assessment in plants is a key to nutrient management. In this research, a nitrogen sensor for corn was developed. With the nitrogen sensor developed, fertilizers could be applied only where it is needed with accurate amount. A lot of expenses would be saved and environmental concerns would be greatly reduced. From the last year’s result (Lee et al, 1999), it was found that the spectral difference resulted from different canopy colors was not affected by variety but by nitrogen content itself in the leaves. Continuing from the last year’s effort, the nitrogen sensor was constructed and tested in a commercial corn field, along with laboratory analysis for N content in sample leaves. The overall objective of this research is to develop a real-time multispectral sensor that could detect nitrogen deficiency in corn plant using spectral response from plant canopies. Eventually the Nitrogen sensor could be used for variable rate fertilizer application. The specific objectives were 1. to construct an in-field sensor system for spectral measurement, 2. to build data acquisition system for real-time field use, 3. to build an algorithm to assess nitrogen status for corn, and 4. to test the N-Sensor in commercial cornfield.