This paper presents the design of an integrated operation machine for growing and picking eggplant in a greenhouse, with the aim of addressing the issues of low efficiency and high cost associated with conventional eggplant management and picking methods. The operating machine incorporates an XYZ-type robotic arm, a computer vision recognition system, an environmental monitoring module, and a 5G communication module. Real-time eggplant recognition is achieved using the YOLOv5 deep learning model, while precise harvesting is facilitated by the XYZ-type robotic arm. Additionally, the system incorporates functions for environmental monitoring and pesticide spraying, thereby enabling intelligent management of the greenhouse environment. Experimental results demonstrate that the system performs exceptionally well in terms of eggplant identification, picking accuracy, and pesticide spraying effectiveness. These advancements significantly enhance the level of automation and overall efficiency in eggplant cultivation, presenting promising application prospects.