High-precision power grid unmanned aerial vehicle inspection image defect identification method and system

The invention discloses a high-precision power grid unmanned aerial vehicle inspection image defect identification method and system, and the method comprises the steps: 1, enabling an unmanned aerial vehicle to fly to shoot and upload flight image data, and 2, adjusting the data of the unmanned aerial vehicle according to the flight image, and obtaining a more precise inspection target image. And step 3, carrying out noise reduction processing on the obtained inspection target image through a Kalman filtering method in combination with a multi-threshold denoising method of a BayesShrink threshold and a Windows Shrink threshold, and carrying out feature extraction processing on the image based on a wavelet packet transformation and mathematical morphology method. 4, comparing and judging the processed image with a fault model, and outputting an identification result; and 5, arranging maintenance after classified statistics. According to the method, the image acquired by the unmanned aerial vehicle is subjecte