Abstract: Abstract: Floor rearing natural mating technique can fully make use of the biological characteristics and animal instincts, in order to exert the genetic advantages and improve the quality of duck eggs. At present, only the pedigree breeding method, rather than the individual breeding method, can be used for floor rearing ducks. The reason is that under the floor rearing condition laying ducks has a large area to move around, and the time of laying duck egg concentrates between 1:00-3:00 in the morning, which makes it extremely difficult to track, identify and record individual laying behaviors and to accurately correlate an individual laying duck with its own eggs manually. The shortcoming of the traditional pedigree breeding method lies in complicated process, high intensity work, low accuracy, long intergenerational interval and slow genetic progress. As the performance of laying ducks is evaluated based on the overall family characteristics, it is hard to tell the difference between high yield individuals and low yield individuals, which negatively affects the precision and efficiency of duck selection. Cage breeding is one of the ways to achieve individual breeding, but this method is subject to a limited activity space and sacrifices the animal's welfare needs. Thus, ducks under such conditions cannot properly exert their biological characteristics and instincts, and the genetic advantage is weakened. Floor rearing provides activity space for laying ducks and effective exercises can enhance their physical quality. Therefore, the realization of the individual laying duck breeding method can not only fully exert the individual instincts but also take the amount of egg production as a key indicator to the future generations. A key to apply the individual breeding method is to find an intelligent and non-stress reaction method to accurately identify and label individual ducks and their duck eggs. For the purpose to investigate laying ducks, a new design framework for an intelligent collection and labeling system was proposed, in which the network topology and a logic control algorithm for both upper monitor and lower monitor were given. This system used the radio frequency (RF) technology and photoelectric sensor technology for information integration, which solved the label missing problem due to the same frequency interference when the RF readers were close to one another. In this way, an accurate identification of individual duck egg production behaviors was achieved. The non-contact inkjet printing technology was employed to label the laying duck identification on its egg shell correspondingly, which solved the problem of recording duck eggs correspondingly. Meanwhile, a new duck egg collection device was designed and implemented; it integrated the collection and marking function as a whole. This device consists of three modules, a collection module, a posture adjustment module, and a printing module. The collection module was designed with a Trapezoidal Groove structure and a flexible contact surface, which eliminated the kinetic energy of duck eggs accumulated in the process of accelerating rolling motion. Such design could reduce the duck egg breakage rate to below 1%. The posture adjustment module was designed with a touch-type duck egg guide. The coincidence rate of the long axis of the duck egg and the parallel axis of the conveyor belt is 99.80%, which ensure that the long axis section of the egg shell was used as the printing surface to maintain the integrity of the printed characters to the best extent. The printing module used a continuous inject printer to mark number on the duck egg and the clearly readable qualified rate of the number marked on the duck egg is 98.2%. The application of the intelligent duck egg collection and labeling system in the field of laying duck breeding will greatly reduce the workload and improve the production efficiency, so as to enhance the production performance and economic benefit of the duck breeding industry.