The fatigue work of air traffic controllers is bound to seriously threaten the air traffic safety. It is a main way to detect the fatigue of air traffic controllers by judging the state of eyes open and closed. In order to detect the controller's fatigue state, a DCNN eye state recognition model based on transfer learning is proposed. Firstly, the MTCNN algorithm of deep cascaded neural network is used to detect the controller's face area. The five key points of the face are calibrated and the eyes are located. Secondly, the detected eye images are transferred to the pre trained DCNN eye state classification model for recognizing the open and closed eyes. Finally, the controller's fatigue state is detected combined with PERCLOSE 80 index. The accuracy, loss rate and F1 score of DCNN, VGG16, InceptionV3 and ResNet50 models were compared on ZJU, CEW and ATCE datasets. Experimental results show that on ZJU and CEW datasets, the detection accuracy of DCNN eye state classification model is 97%, compared with VGG16, InceptionV3, ResNet50 models for eye state classification tasks, DCNN model F1 score has been improved by 3% to 7%. The accuracy of DCNN model detection reached 98.35 percent and the accuracy of F1 score reached 98.06 percent, which verified the validity and accuracy of DCNN model.