Diagnóstico de níveis de nitrogênio em folhas de feijão utilizando visão computacional e redes neurais artificiais

Abstract

Nitrogen nutrition in the bean crop is important to ensure good productivity, however the ways of detecting these levels (chemical and visual analysis) are slow or depend on an experienced professional, so the objective of this work is to diagnose nitrogen levels through of Computational Vision and Artificial Neural Networks (RNA). Beans were grown in a greenhouse containing 5 different doses of nitrogen, 50, 100, 150, 200 and 250 mg L-1. The data collected from the plants were chlorophyll content by means of chlorophyll meter, nitrogen contents in mg L-1 and leaves images, being these realized in two moments of the development of the plant 45 and 58 days after sowing, to make the diagnosis Gray Level Co-Occurrence Matrix (GLCM), non-texturing Statistic and Local Binary Pattern (LBP) were used. Finally, the data generated by the previous methods were used for the training and testing of Artificial Neural Networks (ANN) Multilayer Perceptron for regression and later classification of the levels of nitrogen. This work demonstrated that the three methods are promising depending on the situation, but the combination of the three methods together with a selection of attributes gives better results in the diagnosis of nitrogen.