The invention provides an insulator self-explosion defect detection method based on an inspection image. The insulator self-explosion defect detection method based on the inspection image comprises the steps of acquiring that a to-be-detected inspection image which is shot by an electric power line inspection unmanned aerial vehicle and comprises a glass insulator is obtained, and manually markinga small number of insulator self-explosion defects are manually marked; gGenerating virtual inspection images containing insulator defects in batches in a simulation system; P; performing filtering enhancement processing on the real image and the virtual image to obtain an inspection image joint training set after filtering enhancement; T, training the Faster R-CNN-based deep learning network byusing the joint training set to obtain an insulator self-destruction defect detection modelraining based on Faster R-by using joint training set The CNN deep learning network obtains an insulator self-explosion defect detection model; and the detection model can detect the filtered and enhanced inspection image to be detected, and determine whether the inspection image comprising the glass insulator comprises a self-explosion defect or not. According to the method provided by the invention, the problem of insufficient quantity of big data training sets in previous deep learning training is solved; and the detection accuracy of the self-explosion defect of the insulator is improved.