CHICKEN EGG SEXING BY USING DATA MINING PROCESS


Toksoz C., Albayrak M. , Yasar H.

FRESENIUS ENVIRONMENTAL BULLETIN, vol.30, no.2, pp.1373-1381, 2021 (Journal Indexed in SCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 30 Issue: 2
  • Publication Date: 2021
  • Title of Journal : FRESENIUS ENVIRONMENTAL BULLETIN
  • Page Numbers: pp.1373-1381

Abstract

In this study, relationship between shape index and sex has been explored by data mining processes. This study is intended to eliminate problems such as workload and rooster stowage during sex determination phases after chicken eggs are hatched. In the implementation phase of study eggs are used that obtained from ATAK-S hybrid parents raised in Ankara Poultry Research Institute. 262 eggs in total are recorded after the ambient temperature range is set between 20oC - 24oC. All the eggs are numbered after the shooting with an industrial camera and all the sizes are measured. Then all the numbered eggs are placed in an egg incubator in the poultry farm at Suleyman Demirel University, Agricultural Research and Application Center. Sexes of the hybrid chicks are determined by considering their ATAK-S features after the 21 days of incubation period. Record numbers and sexes of chicks are matched and saved in database. Shape index are estimated by image processing techniques after recordings and size measurements. The data mining processes are initiated after the calculations are converted to proper data format. At the end of data mining process, Random Forest Algorithm had classified male chicks by 93% and female chicks by 100% accuracy rate.