Aiming at the problems existing in the defect detection process of automobile air filter element,a set of online detection scheme for automobile air filter element sealing based on machine vision was proposed and designed.The functions of automatic loading and unloading of filter element,motion control,image acquisition and display,image analysis and processing were realized.In this system,the intermittent movement of conveyor belt,the automatic loading and unloading of parallel robot and the movement of rectangular coordinate robot were controlled by PLC,and the filter image was captured by the camera triggered by the industrial computer.In the image processing algorithm,Visual Studio 2015 was used as the platform,and the open source OpenCV database was used to obtain whether each chamber in the filter element was completely sealed or not.These information was transmitted to PLC by the industrial computer,and then the instructions were sent to the parallel robot through PLC to realize the classification of the filter element.The experimental results show that the system has good realtime performance and high accuracy.It has met the requirements of enterprises detection.