Department of Biomechatronics Engineering, National Pingtung University of Science and Technology, Pingtung 91201, Taiwan.
Sensors (Basel). 2020 Nov 19;20(22):6624. doi: 10.3390/s20226624.
Free-range chicken farming allows egg-laying hens to move freely through their environment and perform their natural behavior, including laying her eggs. However, it takes time to gather these eggs manually, giving rise to high labor costs. This study proposes a smart mobile robot for poultry farms that can recognize eggs of two different colors on free-range farms. The robot can also pick up and sort eggs without damaging them. An egg feature extraction method with automatic thresholding is employed to detect both white and brown eggs, and a behavior-based navigation method is applied to allow the robot to reach the eggs while avoiding obstacles. The robot can move towards the position of each egg via visual tracking. Once the egg is within the collection area of the robot, it is gathered, sorted and stored in the tank inside the robot. Experiments are carried out in an outdoor field of size 5 m × 5 m under different climatic conditions, and the results showed that the average egg recognition rate is between 94.7% and 97.6%. The proposed mobile poultry robot is low in production cost and simple in operation. It can provide chicken farmers with automatic egg gathering on free-range farms.
散养的鸡可以在其环境中自由移动并表现其自然行为，包括产蛋。然而，手动收集这些鸡蛋需要时间，因此劳动力成本很高。本研究提出了一种适用于散养农场的智能移动机器人，可以识别散养农场中两种不同颜色的鸡蛋。该机器人还可以捡起并分拣鸡蛋而不会损坏它们。采用具有自动阈值的鸡蛋特征提取方法来检测白色和棕色鸡蛋，应用基于行为的导航方法使机器人能够在避开障碍物的同时到达鸡蛋位置。机器人可以通过视觉跟踪移动到每个鸡蛋的位置。一旦鸡蛋进入机器人的收集区域，它将被收集、分拣并存储在机器人内部的水箱中。在不同的气候条件下，在 5 m×5 m 的户外场地进行了实验，结果表明，鸡蛋的平均识别率在 94.7%到 97.6%之间。所提出的移动家禽机器人生产成本低，操作简单，可以为散养农场的养鸡户提供自动捡蛋服务。
