Development of a computer vision system for non-destructive quality assessment and classification of chicken eggs

Currently, there is a high demand for poultry products. Breeding of poultry on an industrial scale is possible through the use of artificial incubators. An important factor for egg hatchability is a preliminary assessment of their suitability for incubation, and the use of computer vision systems (CVS) will improve the accuracy of this assessment and eliminate the error associated with the so-called "human factor" from the evaluation process, resulting in an increase in their hatchability. The method of assessing the suitability of eggs for incubation by CVS is not only accurate, but also non-invasive. This article describes a CVS for evaluating the quality of chicken eggs. A selection of hardware components and software for the development of the CVS has been made. A control program has been written to implement the interaction of the components of the CVS within its functionality. The functionality of the CVS includes image acquisition and processing, the determination of large and small egg diameters, and the output of information in text form. Within the framework of this work, the operation of the CVS at various positions of the analyzed samples is analyzed. A comparison of large and small diameters of eggs measured mechanically and by means of CVS has been carried out. The average measurement accuracy was 99.24%.