Recognition and separation of newly-hatched poultry chicks by sex require intensive human labor in broiler hatchery operations. Segregation of chicks by gender and grow-out in gender specific management schemes offer significant benefits to the broiler industry. This project will develop an automated vision inspection system to recognize the sex of baby chicks and separate them into different bins. Our preliminary study indicates the promise of the technology. This research will be conducted with three aspects. First, we will determine the optimum optical parameters for imaging and chick feather pattern extraction and develop image pattern recognition algorithms for chick gender recognition. Next, we will investigate chick wing presentation and handling mechanism for viewing and sorting chicks. Finally, we will investigate the electronic and mechanical interface for synchronization and control of motion vision and separation operations. A working prototype will be developed in Phase II
Anticipated Results/Potential Commercial Applications of Research: The anticipated result will be that the automated vision system can recognize chick feather patterns and distinguish chick sexes. It will be a direct and practical application to the poultry industry in chick separation operations currently done manually. With U.S. processing 160 million broilers each week, the result from this study will advance the poultry operations with improved work environment, labor savings, fee conversion efficiency, and cost-effective production.
