Anomaly Detection Method for Connecting Bolts of Generator Rotor Coils Based on Local Enhancement and Regional Characteristics

The real-time detection of the state of the connecting bolt of the rotor coil in the hydro generator unit is of great practical significance for predicting and analyzing the electrical fault of the generator. In this paper, a bolt image detection method based on local enhancement and regional features is proposed. First, the image is locally contrasted in a fixed window to eliminate the problem of uneven image brightness caused by illumination in the visual detection system; Then, the real-time motion blurred image is restored by Wiener filtering, and the bolt shadow extraction algorithm is proposed to recognize the bolt. The experimental results show that the method has high detection efficiency and reliable performance. This method can meet the real-time detection requirements of generator rotor connecting bolts, and has broad application and promotion prospects.