A new application of morphometric variables and image processing to determine day-old chicken sex

Abstract

One of the principal activities in the poultry industry is determine the sex of chickens of one day old. In this paper, we present a non-invasive technique to determine the sex of day-old chickens based on image processing algorithms. The technique analyzes morphometrical attributes from chicken using the slow and rapid growth of primary and secondary feathers patterns and linear discriminant analysis models. Based on the area formed with the superior points of each feather, the technique is capable to determine the sex of day-old chickens with an accuracy of 94.4%, providing a cheap, non-invasive, and high accurate technique that could be implemented onto a dedicated and automated system.