Application of Computer Vision for quality control in frozen mixed berries production: colour calibration issues

Published: 8 September 2013
Abstract Views: 1039
PDF: 490
Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Authors

Computer vision is becoming increasingly important in quality control of many food processes. The appearance properties of food products (colour, texture, shape and size) are, in fact, correlated with organoleptic characteristics and/or the presence of defects. Quality control based on image processing eliminates the subjectivity of human visual inspection, allowing rapid and non-destructive analysis. However, most food matrices show a wide variability in appearance features, therefore robust and customized image elaboration algorithms have to be implemented for each specific product. For this reason, quality control by visual inspection is still rather diffused in several food processes. The case study inspiring this paper concerns the production of frozen mixed berries. Once frozen, different kinds of berries are mixed together, in different amounts, according to a recipe. The correct quantity of each kind of fruit, within a certain tolerance, has to be ensured by producers. Quality control relies on bringing few samples for each production lot (samples of the same weight) and, manually, counting the amount of each species. This operation is tedious, subject to errors, and time consuming, while a computer vision system (CVS) could determine the amount of each kind of berries in a few seconds. This paper discusses the problem of colour calibration of the CVS used for frozen berries mixture evaluation. Images are acquired by a digital camera coupled with a dome lighting system, which gives a homogeneous illumination on the entire visible surface of the berries, and a flat bed scanner. RBG device dependent data are then mapped onto CIELab colorimetric colour space using different transformation operators. The obtained results show that the proposed calibration procedure leads to colour discrepancies comparable or even below the human eyes sensibility.

Dimensions

Altmetric

PlumX Metrics

Downloads

Download data is not yet available.

Citations

Crossref
Scopus
Google Scholar
Europe PMC

How to Cite

Ricauda Aimonino, D., Gay, P. and Comba, L. (2013) “Application of Computer Vision for quality control in frozen mixed berries production: colour calibration issues”, Journal of Agricultural Engineering, 44(s2). doi: 10.4081/jae.2013.296.

Similar Articles

1 2 3 4 5 6 7 8 9 10 > >> 

You may also start an advanced similarity search for this article.