Liquid Leakage Detection Of Power Plant Pipelines Based On Improved Vibe Algorithm

Liquid leakage in power plant pipelines not only wastes resources, but also affects the safe and stable production requirements of power plants. In order to detect the liquid leakage area of power plant pipelines timely and accurately, and meet the requirements of practical application, this paper proposes a method of power plant pipeline liquid leakage detection based on improved ViBe algorithm. According to the characteristics of power plant pipeline liquid leakage, and the traditional ViBe algorithm cannot adapt to dynamic background changes and quickly eliminate ghosting problems in the detection results, this paper chooses to expand the sampling neighborhood of the background model and adopts an adaptive radius threshold determined by the dispersion coefficient. To improve the accuracy of leak detection, and re-determine the leaking and non-leaking areas through Otsu threshold calculation, quickly and effectively suppress the appearance of ghost areas, so as to make the segmented power plant pipeline liquid leakage area more complete.