Unmanned Aerial Vehicles(UAVs) have a wide range of applications in agriculture, logistics, rescue and disaster relief because of their compactness, lightness and flexibility. However, if they are used improperly or mismanaged, they may not only cause personal privacy leakage and property loss, but also pose a threat to public safety and even military security. Therefore, real-time and accurate detection and warning of UAVs in the airspace play an important role. In this regard, a multi-channel interactive attention mechanism and edge contour enhancement (MCIAECE) method for infrared UAV detection is proposed. Firstly, the shallow and deep features of the infrared image are extracted by a dual-channel consisting of a multi-channel interactive attention mechanism module and an edge contour enhancement module, after which the attention mechanism enhances the target features while the edge contour enhancement obtains more detailed information. Then the extracted features of each layer are fused and enhanced using the multilevel feature fusion module to obtain the detection results. The experimental results show that better results can be achieved with the MCIAECE method on all three datasets. Among them, the best results are obtained on the NUDT-SIRST infrared dataset, with the detection probability and intersection over union of 98.83% and 85.11% respectively, which increased by 1.95% and 6.88% compared to the baseline network,and the effect is significant in the edge contour restoration of the target compared with other methods.