As the key technology of transmission line inspection, insulator detection plays an important role in maintaining the safe and stable operation of transmission system. In order to overcome the shortcomings of existing methods, such as easily losing target location information and low precision of insulator detection in complex background, an insulator detection method based on feature pyramid and multi-task learning is proposed, which combines high and low dimension feature information to construct feature pyramid and avoid the loss of detail information such as target location and realize efficient detection of insulators in complex background. The experimental results show that the proposed method can improve the detection accuracy of insulators to 95.3%, and has high engineering application value.