Aiming at the problem of inability to accurately detect pipeline liquid dripping in coal mine

a liquid dripping automatic recognition method based on inter-frame difference calculation is proposed.Unmanned vehicles inspect underground environments by autonomous navigation

using cameras to capture videos

using YOLOv5 model to locate locations liquid leakage may occur

and then

using inter-frame difference and Hough transform methods to judge whether liquid dripping has occurred or not.If liquid dripping occurs

the location of the leakage is marked and transmitted to the software platform

and coal mine personnel are notified to handle it in a timely manner.The results show that the recognition rate of the system can reach 98%

with a fast recognition speed and without distinguishing liquid types

making it suitable for hazardous areas such as mine shafts.