Autor(es): Guedes, Wesley Nascimento [UNESP] ; dos Santos, Lucas Janoni [UNESP] ; Filletti, Érica Regina [UNESP] ; Pereira, Fabíola Manhas Verbi [UNESP]
Data: 2019
Identificador Persistente: http://hdl.handle.net/11449/187746
Origem: Oasisbr
Assunto(s): Artificial neural networks; Digital images; Food quality; Sugar
Made available in DSpace on 2019-10-06T15:45:59Z (GMT). No. of bitstreams: 0 Previous issue date: 2019-01-01
For the first time in literature, an analytical method was developed using artificial neural networks (ANNs) combined with color information from digital images to predict the content of sugarcane stalks in the presence of a solid impurity. The data were generated using a laboratory-made simple imaging system and free-access computational routine for the conversion of the images into 10 colors. The ANN model was implemented using 10 neurons in the input layer, 8 neurons in the hidden layer and 1 neuron in the output layer related to the content of sugarcane stalks. The ANN model provided relative errors of 3% and achieved correlation coefficients of 0.98, 0.93, and 0.91 for the training, validation and test sets, respectively. A partial least squares (PLS) model showed the nonlinear nature of the data that implies the application of ANN model. The developed method has the potential to be applied in sugarcane mills as an improvement for the production of high-quality sugar.
Bioenergy Research Institute (IPBEN) Institute of Chemistry São Paulo State University (UNESP)
Bioenergy Research Institute (IPBEN) Institute of Chemistry São Paulo State University (UNESP)
