The product quality and cost of Continuous Vertical Galvanizing Line (CVGL) is seriously affected by strip vibration. Detection and reduction of the vibration is focused in this area. Aiming at the CVGL strip vibration detection problem, this paper proposes a strip overall vibration detection and analysis method based on collaborative vision of multi channel.First sample sections that show strip overall vibration are selected,5-7 key points are determined in each section according to precision required. The vibration of key points is measures by monocular vision, ie, the vibration is calculated through the five laser flare as strip vibration in the camera imaging element. The video signal will carry strip vibration. The information of multi-channel monocular visual with timestamp bounded summary to the master computer. The vibration amplitude, frequency and other information could be found through the analysis of the image of each section. Finally, the overall vibration of strip could be found through multi-channels collaborative analysis of the visual information.The timestamp consistency of each channel is guaranteed by the synchronous broadcasting of synchronizing signal on RS485 bus. In order to eliminate the influence of Ethernet propagation delay, and polymorphism join synchronous serial number in every video packets, so to achieve a wide range of acquisition synchronization. Filed test shows that the scheme is feasible. |