Abstract: A machine vision-based human facemask wearing detection device was developed to realize the visual recognition of facemask wearing. The device is composed of two parts: the lower computer part and the host computer part. The former uses the embedded Linux system as the bottom platform of the system, adopts the V4L2 video capture driver framework, and uses the USB camera to capture the image in real time, and the collected image is transmitted to the host computer through Socket network communication. The latter introduces the OpenCV Machine Vision Open Source Library, and uses the cascade classifier to detect whether the human facemask is worn or not. The experimental results show that the device can effectively solve the preset problems, and integrates machine vision, embedded application technology and other aspects of knowledge to provide ideas and solutions for the design of other automatic control experimental instruments, which has high practicality, application and promotion value.