Application of image analysis to the prediction of EBC barley kernel weight dis...

• Autor(es): Amaral, A. L. ; Rocha, Orlando ; Gonçalves, Cristina ; Ferreira, António Augusto ; Ferreira, E. C.
• Data: 2009
• Identificador Persistente: http://hdl.handle.net/1822/9613
• Origem: RepositóriUM - Universidade do Minho
• Assunto(s): Barley analysis; Kernel size; Image analysis; Partial least squares
• Tipo de Documento: Artigo
• Idioma: Inglês

Descrição

It is known that the barley kernel size is an important factor regarding the uniformity of the malting and brewery processes, barley valuation, approval andmarket value. In order to facilitate the barley purchasing process, a fast field technique for kernel size evaluation, such as the image analysis technique proposed in this work, would be greatly appreciated as a fast and simple procedure for barley selection. In this study a close correlation between the image analysis and the standard EBC was obtained with a correlation factor of 0.999 and a regression coefficient of 0.991 between the two methodologies. The proposed IA methodologywas found to accurately predict the Scarlett and Prestige barley varietiesweight distribution especially when considering the crucial ‘business transactions selection’ classes.