AIRPORT FLIGHT ZONE REAL-TIME TARGET DETECTION METHOD BASED ON MULTISCALE FEATURE DECOUPLING

Disclosed is an airport flight zone real-time target detection method based on multi-scale feature decoupling, the steps of which comprise: 1. Acquiring airport flight zone monitoring video, and constructing an airport flight zone target detection dataset; 2. Constructing a multiscale feature fusion module so as to detect targets of different scales; 3. Decoupling classification tasks and positioning tasks within target detection, and constructing a learning network based on feature decoupling; 4. Adding to a YOLOv5 target detection network the learning network based on feature decoupling and the multiscale feature fusion module, and combining with a loss function to train and optimize a target detection model. The present invention utilizes a multiscale feature fusion module to aggregate shallow detail information and deep semantic information, thereby enhancing capabilities for detecting targets of different scales, and utilizes a feature decoupling learning network to decouple classification tasks and posi