The system is made up of amachine vision devicewhich employs an algorithm based on a combination of digital images taken at appropriate wavelengths – red and blue. Through the proposed algorithm, the images of the analyzed samples -384- were segmented, obtaining, in the case of eggs with a defective shell, a binary image capable of showing the defects present on the surface of the egg.
Based on the geometric characteristics of the spots detected, the system achieves ccorrectly lasify 98% of samples with very low processing time: only 0.05 seconds.
Since the invention uses an algorithm based on two wavelengths (red and blue) that, with the green one, constitute the RGB filter of a common digital camera, this method constitutes a simple, fast, cheap, and non-destructive technique for the automatic grading of eggs for consumption and could be considered as an important first step towards the automation of the entire process.
In Europe, the production process of poultry eggs intended for human consumption has three stages: collection, classification and packaging. Although in recent years the harvesting and packaging phase have been largely automated, the grading phase, during which the eggs are graded and inspected for any defects, is still done manually.