The appropriate amount of antioxidant compounds added in frozen light-flesh cut fruit production process is usually estimate evaluating colour changes, in particular by measuring the minimum time before browning phenomena occur. This parameter is assessed by visual inspections performed by trained operators: a time consuming and strongly subjective procedure. The development of a Computer Vision System (CVS) for quality control in frozen fruit slice is presented in this paper. An algorithm to detect and measure browned area on fruit slices was implemented in order to describe colour changes evolution. The red component of browned areas is emphasized by an adequate linear combination of RGB colour channels of digital images, and an entropy-based automatic segmentation is applied to the obtained high contrast grey-scale image. This approach is not based on colour measurement of samples by the CVS, avoiding the colour calibration phase. Antioxidant solutions with different concentrations were applied in order to obtain different browning times and evaluate algorithm performances. Results obtained with the CVS strongly fit with visual inspections performed by trained operators, showing the reliability of the method for this specific application.

Computer vision for laboratory quality control on frozen fruit

RICAUDA AIMONINO, Davide
First
;
BARGE, Paolo;COMBA, Lorenzo;GAY, Paolo;OCCELLI, ALESSANDRO;TORTIA, Cristina
2015-01-01

Abstract

The appropriate amount of antioxidant compounds added in frozen light-flesh cut fruit production process is usually estimate evaluating colour changes, in particular by measuring the minimum time before browning phenomena occur. This parameter is assessed by visual inspections performed by trained operators: a time consuming and strongly subjective procedure. The development of a Computer Vision System (CVS) for quality control in frozen fruit slice is presented in this paper. An algorithm to detect and measure browned area on fruit slices was implemented in order to describe colour changes evolution. The red component of browned areas is emphasized by an adequate linear combination of RGB colour channels of digital images, and an entropy-based automatic segmentation is applied to the obtained high contrast grey-scale image. This approach is not based on colour measurement of samples by the CVS, avoiding the colour calibration phase. Antioxidant solutions with different concentrations were applied in order to obtain different browning times and evaluate algorithm performances. Results obtained with the CVS strongly fit with visual inspections performed by trained operators, showing the reliability of the method for this specific application.
2015
44
175
180
http://www.aidic.it/cet/15/44/030.pdf
Chemical Engineering (all), fruit, vision
RICAUDA AIMONINO, Davide; Barge, Paolo; Comba, Lorenzo; Gay, Paolo; Occelli, Alessandro; Tortia, Cristina
File in questo prodotto:
File Dimensione Formato  
Chemical_Engineering_Transactions_2015.pdf

Accesso aperto

Tipo di file: PDF EDITORIALE
Dimensione 741.18 kB
Formato Adobe PDF
741.18 kB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2318/1564400
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 4
  • ???jsp.display-item.citation.isi??? 2
social impact