The microinjection of Zebrafish embryos is significant to life science and biomedical research. In this article, a novel automated system is developed for cell microinjection. A sophisticated microfluidic chip is designed to transport, hold, and inject cells continuously. For the first time, a microinjector with microforce perception is proposed and integrated within the enclosed microfluidic chip to judge whether cells have been successfully punctured. The deep learning model is employed to detect the yolk center of zebrafish embryos and locate the position of the injection needle within the yolk, which enables enhancing the precision of cell injection. A prototype is fabricated to achieve automatic batch microinjection. Experimental results demonstrated that the injection efficiency is about 20 seconds per cell. Cell puncture success rate and cell survival rate are 100% and 84%, respectively. Compared to manual operation, this proposed system improves cell operation efficiency and cell survival rate. The proposed microinjection system has the potential to greatly reduce the workload of the experimenters and shorten the relevant study period.