Automatic Detection Method of Operation Error of Smart Electricity Meter Based on Machine Vision

In order to improve the detection effect and reduce the number of electricity theft, an automatic detection method of operation error of intelligent electricity meter based on machine vision is proposed. Firstly, the smart meter image is collected, the Otsu threshold binarization method is used to extract the pulse lamp area image, and the improved lifting wavelet and morphology are used to preprocess the area image. Then, the pulse lamp in the reconstructed image is segmented by projection combined with a priori knowledge feedback method, and the template matching method is used to automatically detect whether the segmented pulse lamp is on, Finally, the experimental results show that this method can effectively detect the meter operation error under different line loss and power factor, effectively reduce the occurrence times of power theft, and has high detection accuracy. When different image brightness, this method has high dice value and high segmentation effect quality.