Based on the embedded ARM-Linux platform and OpenCV open source visual library, a online detection device for the CNC tool wear status was designed. In the system, the video module V4L2 of Linux was adopted to carry on the data acquisition from the industrial camera, the OpenCV intelligent algorithm and the contour processing correlation function were used to extract the tool wear characteristic value (VB) in the embedded image acquisition side, wear status was judged by the VB. In addition, through the terminal WiFi module AP function, the image data were transmitted to other network devices for display, to achieve wireless transmission capabilities. The test results show that the system achieves the desired goal.