The performance of the seedling picking device, as a core component of transplanting machines, is directly related to the efficiency, quality, and cost of the transplanting operation, and plays an active role in promoting the development of automatic transplanters. At present, a contradiction exists between the seedling picking speed and the success rate of seedling picking. With the increase in the vegetable planting area, the quality requirements for seedling picking are gradually increasing. Efficient and high-quality seedling picking has become the focus of scholars. In order to provide scholars with a comprehensive and profound understanding of seedling picking devices, this study searched and screened the literature, and the CiteSpace literature quantitative analysis software was employed to analyze the research hotspots and trends by using a knowledge graph. Three kinds of seedling picking devices and pot seedling detection technologies were discussed. The working principles, advantages, and disadvantages of these seedling picking devices and detection technologies were analyzed, and the problems with current seedling picking devices were summarized. Results show that, the automatic seedling picking device has a complex structure, weak versatility, difficulty in fully matching diversified agronomic needs, low intelligence level, large research and development investment, high cost, and difficulty in promotion. Solutions are given for these corresponding problems: the seedling picking device should adopt a modular design to improve its versatility; the creation of unified standards for the production of pots, the growing of seedlings, and the development of sensors must be accelerated; and computer vision, artificial intelligence, deep learning, intelligent control algorithms, and other technologies must be integrated into seedling picking devices to improve their level of intelligence, ensure transplanting efficiency, and reduce damage to pot seedlings. A universal high-speed, low-loss seedling picking device should be the main research direction in the future. This study provides a good reference for the further development of automatic transplanting machines.