Abstract: In order to solve the safety problem of micro-cultivator, a safety protection system based on attitude recognition and internet of things alarm was developed. Firstly, the overall design scheme and functional principle of the safety system of the micro-cultivator was described, that is, the fall event of human body can be identified by the attitude data of the micro-cultivator fed back by the gyro acceleration sensor, and the micro-cultivator would be shut down before the danger occurs. At the same time, remote alarm can be implemented through the internet of things system, which can effectively solve the secondary safety protection problem when the onboard controller fails. Secondly, the hardware and software of the security system were introduced, including the specific design scheme of the onboard controller and the Internet of Things system. Finally, bench tests and field tests were performed for the onboard controller, and remote communication tests were performed for the internet of things system, which verifies the effectiveness and rationality of the system. The results show that the safety system is not only easy to install, safe and reliable, but also has low application cost, which can meet the safety protection needs of the micro-cultivator under general working conditions, and has the value of further application.