Development of Infrared Sensor Based Data Acquistion System for Predicting Soil Nitrogen Content as a Reference to Determine Fertilizing Dosage

Abstract

Site specific nitrogen fertilizing needs an accurate map of soil nitrogen content. The use of sensors operated upon the soil is a promising method since the accurate soil sampling methods are costly and time consuming. The objectives of this research are to determine the relation between soil nitrogen level and near infrared spectrum using artificial neural network (ANN) and to develop soil nitrogen content data acquisition system for static dan dynamic measurement. The results showed that the 1506 nm wavelength can be used to estimate the soil nitrogen content. Furthermore it was found that static measurement showed a better correlation (R2= 0.6286) than the dynamic measurement (R2=0.3111). Combined with the developed ATMega32 microcontroller based display recorder, the precision of N content measurement achieved 0.12% wb with 0.1% wb noise.